Victorian Cardiac Outcomes Registry ANNUAL PUBLIC REPORT 2016 Improving cardio vascular outcomes Victoria-wide

This publication was produced on behalf of the Victorian Cardiac Outcomes Registry (VCOR).

Suggested citation: A/Prof Jeffrey Lefkovits, Ms Angela Brennan, Dr Diem Dinh, Ms Rita Brien, Mrs Harriet Carruthers, Mrs Janine Doyle, Dr Dion Stub, Prof Andrea Driscoll and Prof Chris Reid on behalf of the VCOR. The Victorian Cardiac Outcomes Registry Annual Report 2016. Monash University, DEPM June 2017, Report No 4, 63 pages.

Any enquires or comments regarding this publication should be directed to:

VCOR C/- Department of Epidemiology & Preventive Medicine Monash University Commercial Road VIC 3004

Phone: +61 3 9903 0302 Email: [email protected]

Report No: 4, June 2017 Contents VCOR 2016 Annual Report

4 List of Figures 5 List of Tables 6 Acknowledgements 7 Foreword 8 Executive Summary 9 Key Findings 9 PCI Registry 10 Management of Acute STEMI in Rural and Regional Centres 10 Heart Failure Snapshot 11 Introduction 11 Registry Governance and Structure 12 Registry Activities 13 Percutaneous Coronary Intervention (PCI) 13 Registry Module Activity 14 Data Quality - Audit Activity 15 Patient Characteristics 16 Clinical Presentation 18 Indications for PCI 20 Clinical and Lesion Subsets 21 Device Use 22 Drug Eluting Stents 24 Arterial Access 25 PCI for Acute STEMI 27 Door-to-balloon Times for PCI for Acute STEMI (Primary PCI) 29 Outcomes 29 Lesion and Procedure Success Rates 29 New Renal Impairment 30 Referrals to Cardiac Rehabilitation 31 Compliance with Guideline-Recommended Discharge Medications 32 Key Performance Indicators 38 30-day stent thrombosis 38 30-day rehospitalisation 39 Quality of Life Metrics 40 Management of Acute ST-Elevation Myocardial Infarction (STEMI) in Regional Victoria (Early STEMI Management) 40 Background 40 Registry Module Activity 40 Data Linkage with VCOR STEMI module 41 Patient Characteristics 42 Time Delays to Transfer 45 Adjunctive Therapies 45 In-hospital Outcomes and Transfer Rates 45 Revascularisation Rates 48 Heart Failure (HF) Snapshot 48 Background 48 Registry Module Activity 49 Patient Characteristics 49 Clinical Presentation 53 Transitional Care after Discharge 54 Outcome measures 56 Future Directions 57 Glossary 58 Publications and presentations in 2016 59 VCOR Personnel 61 Funding 62 References List of Figures

14 Figure 1: Cumulative cases submitted by month from 34 Figure 29: Rates of in-hospital major bleeding 2013-2016 34 Figure 30: Rates of in-hospital unplanned revascularisation 14 Figure 2: Overall rate of missing cases across each annual 35 Figure 31: Length of stay by clinical presentation audit (2014-2016) 36 Figure 32: Risk-adjusted 30-day mortality 15 Figure 3: Age and gender distribution of patients 36 Figure 33: 30-day mortality rates for cardiogenic shock undergoing PCI and intubated OHCA patients by site 16 Figure 4: Age distribution for public and private patients 37 Figure 34: 30-day MACCE 16 Figure 5: Procedures by clinical presentation 35 Figure 35: 30-day unplanned cardiac rehospitalisation by 17 Figure 6: ACS and non-ACS cases by hospital hospital 17 Figure 7: Procedures by clinical presentation for public 39 Figure 36: EQ-5D responses to anxiety/depression by age and private hospitals group 19 Figure 8: Clinical indicators for PCI in non-ACS patients for 42 Figure 37: Ambulance time intervals for regional STEMI public and private hospitals patients 20 Figure 9: Comparative trends in incidence over time for 42 Figure 38: Time from arrival to first ECG time for regional selected PCI lesion subsets STEMI patients by hospital 21 Figure 10: Device use among PCI cases 43 Figure 39: Door-to-needle times for regional STEMI 21 Figure 11: Number of stents used per PCI case (includes all patients by hospital lesions treated) 43 Figure 40: Proportion achieving door-to-needle times 21 Figure 12: Total length of stents used per PCI case within 30 and 60 mins (regional STEMI patients) (includes all lesions treated) 44 Figure 41: Overall system delay times (regional STEMI 22 Figure 13: DES use by hospital patients) 23 Figure 14: Trends in DES usage by hospital sector: 2013- 45 Figure 42: Treatment and outcomes: adjunctive therapies 2016 (regional STEMI patients) 23 Figure 15: Device use by clinical presentation 46 Figure 43: Patient transfer times to metro VCOR hospital 24 Figure 16: Trends in arterial access: 2013-2016 (regional STEMI patients) 24 Figure 17: Arterial access route by hospital 46 Figure 44: Regional STEMI module inter-hospital transfer 25 Figure 18: Arterial access route in public and private and revascularisation rates hospitals 50 Figure 45: New York Heart Association (NYHA) class rates 25 Figure 19: Acute STEMI cases as a proportion of overall on admission and discharge during HF-Snapshot case numbers by hospital 51 Figure 46: Prescribing of beta-adrenergic blockers in HFrEF 26 Figure 20: PCI treatment type for acute STEMI patients by cohort with heart rate >60 BPM on admission and hospital discharge during HF-Snapshot 26 Figure 21: Radial access rates in acute STEMI cohort 52 Figure 47: Prescribing of ACEI/ARB in HFrEF with eGFR >60 on admission and discharge during HF-Snapshot 27 Figure 22: Door-to-balloon time for primary PCI cases by hospital 53 Figure 48: In-hospital all-cause mortality during HF- Snapshot 28 Figure 23: Proportion of primary PCI cases with door-to- balloon time ≤90 minutes by hospital 53 Figure 49: Discharge planning during HF-Snapshot 28 Figure 24: Proportion of primary PCI cases with door-to- 54 Figure 50: Overall outcome measures during HF-Snapshot balloon time ≤90 minutes - pre-hospital notification (PHN) 2016 vs no pre-hospital notification 54 Figure 51: Unadjusted 30-day mortality during HF- 29 Figure 25: Proportion of primary PCI cases with door-to- Snapshot 2016 balloon time ≤90 minutes - in-hours vs out-of-hours 55 Figure 52: 30-day all-cause hospital readmission during presentation HF-Snapshot 2016 30 Figure 26: Rate of new renal impairment in selected high- 55 Figure 53: Time to readmission by hospital for HF- risk subgroups Snapshot 2016 30 Figure 27: New renal impairment rates by clinical presentation 33 Figure 28: Comparative in-hospital mortality trends for different clinical presentation groups (2013-2016)

4 VCOR | 2016 Annual Report List of Tables

13 Table 1: Participation of Victorian PCI hospitals 15 Table 2: Selected patient characteristics 2013-2016 16 Table 3: Selected patient characteristics by hospital sector 18 Table 4: PCI Indication by ACS category and sub-category 18 Table 5: PCI Indications for non-ACS cases 19 Table 6: Non-ACS patients: Clinical Indicators for PCI 20 Table 7: Patients presenting with cardiogenic shock or out-of-hospital cardiac arrest (OHCA): 2013-2016 22 Table 8: Adjunctive device use 27 Table 9: Door-to-balloon-time for primary PCI cases 31 Table 10: Referral rates to cardiac rehabilitation by clinical presentation 31 Table 11: Discharge Prescription Rates of Beta Blockers (BB) by clinical presentation 31 Table 12: Discharge Prescription Rates of ACE Inhibitor/Angiotensin Receptor Blockers (ARB) by clinical presentation 32 Table 13: Unadjusted in-hospital mortality rates for selected patient groups 33 Table 14: In-hospital major bleeding rates by sub-group 37 Table 15: 30-day MACCE rates 38 Table 16: Rehospitalisation rates 39 Table 17: Quality of life metrics for 2016 PCI cohort 41 Table 18: Participation of regional Victorian hospitals in Early STEMI Management module 41 Table 19: Regional STEMI patient characteristics (2013-2016) 44 Table 20: Door-to-needle times for ambulance pre-hospital notification (PHN) and no PHN (regional STEMI patients) 45 Table 21: In-hospital outcomes for regional STEMI patients 47 Table 22: Outcomes for regional STEMI patients by reperfusion strategy 48 Table 23: Participation of hospitals in HF-Snapshot (2014-2016) 49 Table 24: HF-Snapshot patient characteristics (2015-2016) 50 Table 25: HF-Snapshot clinical presentation (2015-2016) 51 Table 26: Medications prescribed at admission and discharge for all patients during HF-Snapshot 52 Table 27: Discharge destination during HF-Snapshot 61 Table 28: VCOR Funding 2011 – 2016

VCOR | 2016 Annual Report 5 Acknowledgements

VCOR is funded by the Department of Health and Human Services (DHHS) Victoria with in-kind funding from Monash University. VCOR is operated by the Department of Epidemiology and Preventive Medicine (DEPM), Monash University, in association with the Victorian Cardiac Clinical Network, DHHS Victoria.

We gratefully acknowledge contributions made by the VCOR In 2016 the Registry Custodian, Prof Chris Reid was Steering Committee, the VCOR Clinical Quality Committee, supported by a National Health and Medical Research the VCOR Data Research and Publications Committee, the Council (NHMRC) Fellowship that provided salary support to VCOR Registry Custodian - Prof Chris Reid and the VCOR contribute to initiatives such as the VCOR. Clinical Director - A/Prof Jeff Lefkovits. We also acknowledge the VCOR Data and Project Management Committee staff at This report would not have been possible without the efforts DEPM (Ms Angela Brennan, Dr Diem Dinh, Dr Nick of doctors, nurses, data managers and other relevant Andrianopoulos, Dr Dion Stub, Prof Andrea Driscoll, Ms Rita hospital staff who contribute data to the VCOR. Brien, Ms Harriet Carruthers, Mrs Janine Doyle, Mr Mark Lead clinical staff from the hospitals participating in the VCOR Tacey and Mrs Jill Edmonds). are also gratefully acknowledged.

6 VCOR | 2016 Annual Report Foreword

As the inaugural Chief Executive of Safer Care Victoria, it is The future of health outcome reporting will see increasing my great pleasure to introduce the Victorian Cardiac use of the patient voice, such as is evident in the standard Outcomes Registry (VCOR) Annual Report for 2016 – the sets developed by the International Consortium for Health fourth such report from VCOR. Outcomes Measurement (ICHOM) and as presented in the VCOR report. This is important. At the end of the day – Our mission at Safer Care Victoria is “Outstanding health indeed at the beginning of the day – we are here to meet to care for all Victorians. Always”. Working with health services needs of our consumers – patients, their families and carers. and health practitioners to deliver on that mission, and to Asking them what is important is crucial to delivering on our show that, together, we are delivering on that mission, mission. Safer Care Victoria looks forward to working with requires us to measure, analyse and report outcomes. all of our clinical quality registries, our clinical networks, Without meaningful data, delivered in a timely manner into clinicians and health services, and of course consumers to the hands of those to whom the data matter most – provide outstanding health care. Always. VCOR is showing clinicians and consumers – we cannot expect to even begin us the way in cardiac services. on the improvement journey. In the words of Galileo “Measure what can be measured, and make measurable On behalf of Safer Care Victoria, the Department of Health what cannot be measured”. and Human Services, and indeed of all Victorians, I congratulate Associate Professor Lefkovits, the VCOR team With the intent of informing and targeting improvement, and site leads, and all participating clinicians and hospitals clinical quality registries have become an important and on their commitment to improving cardiac services in our effective tool in measuring, analysing and reporting health state. I commend the Annual Report to you and look forward outcomes. The embedded clinical expertise in the design, to next year’s report with even better outcomes. running and reporting of registries is, in my view, the critical component of registry successes. The clinician voice ensures that the measures are relevant and meaningful and that the interpretation of the data is appropriately focussed on genuine opportunities for improvement. Let’s be honest, nothing quite holds clinicians to account than other clinicians! In this way, clinical expertise has ensured that registries deliver on the expectations of the National Health and Hospitals Reform Commission (2009) that we have “systems in place to provide comparative clinical performance data back to health services and … clinicians”.

VCOR and its Annual Reports are an exemplar of such clinical performance reporting. That all Victorian hospitals, public and private, that provide percutaneous coronary interventions now participate in VCOR is a measure of both the success of the registry and of the value that clinicians clearly place in the reported comparative data. In that regard, as you will read, the data include measures of process – time from arrival to first ECG or door-to-needle times, for example – and of outcomes, both clinical and Professor Euan Wallace AM patient reported. A standardized approach to outcome CEO, Safer Care Victoria reporting, such as VCOR reporting, is an essential component of quality improvement and of ensuring the provision of safe and appropriate care. This is core business for Safer Care Victoria. I welcome the VCOR report and look forward to working with our Cardiac Clinical Network to applying its lessons.

VCOR | 2016 Annual Report 7 Executive Summary

In October 2016, the Victorian Department of Health and interest are the increasing number of high-acuity patients Human Services released a review of hospital safety and with cardiogenic shock or out-of-hospital cardiac arrest and quality assurance in Victorian hospitals. The report, Targeting the continuing burden of emergency PCI for acute ST Zero – Supporting the Victorian Hospital System to Eliminate elevation myocardial infarction that is predominantly taken Avoidable Harm and Strengthen Quality of Care[1] was up by the public sector. The Report also reveals changing undertaken after the tragic events at Bacchus Marsh Hospital patterns of practice including widespread uptake of the radial in 2013-2014 that resulted from avoidable and potentially artery for arterial access and near universal usage of drug- avoidable deficiencies in care. Among the report’s many eluting stents. Overall outcomes remained similar to and crucial recommendations, was a strengthened role for clinical often better than international benchmarks. When the quality registries to facilitate the identification of any potential comparison of hospital outcomes identified a performance shortcomings in care. Also in 2016, the Victorian Department outlier, the registry provided timely feedback to the affected of Health and Human Services released its new cardiac health service, facilitating their own internal review of their services plan - Design, service and infrastructure for Victoria’s processes and results. cardiac system[2] . In this report, the utility of registries in ensuring safe and effective care was once again highlighted. In the second section of the Annual Report relating to early management of acute myocardial infarction in non- It is in this setting that the Victorian Cardiac Outcomes metropolitan settings, the number of contributing hospitals Registry (VCOR) has continued its mission to improve cardiac has grown to 9 in 2016, with plans to roll out the program to care across the state. The registry’s core business is to other hospitals across the state in the coming 1-2 years. The measure and report on the quality and safety of patient care Report indicates that while these hospitals remain actively - within individual hospitals, comparatively with other engaged in ongoing service improvement activities, there are hospitals, and aggregated at the state level. Safer Care Victoria still challenges to overcome. Treatment goals relating to now oversees its activities and provides funding through the timely and effective emergency treatment for acute ST Victorian Cardiac Clinical Network. elevation myocardial infarction were not fully met, and a number of sites struggled to maintain resources for reliable This Annual Report presents the activities of VCOR over 2016 data collection and entry. An exciting innovation in 2016 was in three separate areas of interest in cardiovascular care. the establishment of data linkage with Ambulance Victoria for These include percutaneous coronary intervention (PCI), the this patient group with synergistic data exchange benefitting early treatment of acute myocardial infarction in rural and both organisations. regional settings, and data relating to in-hospital management of heart failure. The first two directly relate to management The third section of this Report presents the results of the of coronary artery disease, primarily in its acute form (heart Heart Failure (HF)-Snapshot module. Unlike the other two attacks and angina). The third focuses on chronic heart modules, data for this module were collected in the form of a disease (both coronary and non-coronary) that places a “snapshot”, enrolling consecutive patients at participating significant load on the individual and society in terms of health services for a limited period of time (1 month) in order disability, reduced quality of life, high healthcare resource to obtain a cross-sectional picture of heart failure-related consumption and costs. treatment and outcomes. The number of hospitals participating in this module grew to 16 in 2016, with data In 2016, PCI data were captured on almost 10,000 cases collection activity undertaken in May-June 2016. There were across 13 public and 12 private hospitals. VCOR now has mixed results with generally good compliance rates with cumulative data on over 36,000 PCI cases, collected since its guideline-recommended medical therapies and short-term commencement in 2013. This year also marked the mortality rates comparable to international rates, but achievement of one of VCOR’s main goals with unanimous relatively poor transitional care arrangements. Thirty-day engagement of all hospitals performing PCI in Victoria (public readmission rates for heart failure were lower than for the and private). While 5 private hospitals are yet to commence previous year, but there is still room for further improvement. data contribution, it is envisaged that in 2017, VCOR will be able to report on all PCIs performed in the state. The key findings from the various modules in 2016 are presented on the following pages. The Report follows trends in patient demographics and emerging treatment patterns in PCI. Particular areas of

8 VCOR | 2016 Annual Report Key Findings

PCI Registry

• A total of 25 out of 30 Victorian PCI hospitals contributed • Emergency treatment for acute STEMI (primary PCI) to the registry in 2016, representing approximately 100 accounted for 16% of the overall PCI workload among interventional cardiologists. Data on 9,992 completed VCOR hospitals. The proportion of hospitals’ STEMI procedures involving 8,893 patients were collected. 62% workload that fell out-of-hours was 61%. The majority of cases were managed in the public hospital system. (86%) were treated in the public hospital system. Radial access was used in 60% of cases. • The majority of patients undergoing PCI were male (77%). The mean age of patients was 66 years. Patients treated • The median time taken from patient arrival at the hospital in private hospitals were six years older on average than to the first inflation of the balloon to re-open the artery public patients. Overall, approximately one in five patients (door-to-balloon time) for acute STEMI PCI cases was 67 were diabetic. min, within the recommended threshold of ≤90 minutes. Overall, a door-to-balloon time ≤90 minutes was achieved • Just over half the PCI cases in 2016 presented with an in 72% of cases, close to the international benchmark acute coronary syndrome (ACS). Their treatment was target of 75% or more cases. predominantly taken up by the public sector (77% public sector, 23% private). Treatment of ACS accounted for two- • The unadjusted in-hospital mortality rate overall was thirds of all PCI work done in the public sector and one- 1.8%. The rate was higher among patients presenting with third of the private sector’s caseload. PCI for acute STEMI STEMI (6.1%), and highest for patients with cardiogenic (including pharmaco-invasive PCI and rescue PCI) took up shock or out-of-hospital cardiac arrest requiring 27% of the entire PCI workload in the public sector. endotracheal intubation (37.3%). Excluding these 2 high- risk groups, the unadjusted in-hospital mortality rate for • For patients with stable (non-ACS) disease, 83% had a the rest of the cohort was just 0.4%. diagnosis of stable angina. A high-grade stenosis was noted in 92% and positive functional test in 42%. A • The signature key performance indicator of risk-adjusted total of 87% of non-ACS patients had at least 2 of these 30-day mortality for the overall PCI cohort in 2016 was 3 PCI indicators. 2.8%. One hospital was identified as an outlier with a higher than expected risk-adjusted mortality rate and it • Treatment was predominantly single vessel PCI, with the received appropriate notification and an offer for majority of lesions (66%) between 11-30mm in length guidance and assistance with subsequent quality and the majority of patients (67%) receiving a single assurance activities. stent. Drug-eluting stent use increased from the previous year and was at 85% of cases. Functional • Results of other outcome measures demonstrated an assessment of coronary lesions by fractional flow in-hospital major bleeding rate following PCI of 1.0%, reserve (FFR) was utilised in 2.5% of cases. Other being lower among radial access cases (0.5% radial vs adjunctive devices such as intravascular ultrasound and 1.5% femoral) and highest in STEMI (2.1%). The 30-day rotational atherectomy were performed in around 1% of unplanned cardiac readmission rate was 3.7% for the cases. Bioresorbable vascular scaffolds (BVS) were entire cohort with similar rates in the public and implanted in very small numbers in 2016. private sectors.

• For the first time, radial artery access was utilised more commonly than femoral access, at 55% of cases overall. Rates still varied quite widely among hospitals, with public hospitals generally having a higher uptake at 63% compared with private hospital radial access uptake at 43%. Lower rates were seen in females than males (48% vs 57%) and in the elderly (43% >80 years vs 57% <80 years).

VCOR | 2016 Annual Report 9 Key Findings continued ...

Management of Acute STEMI in Heart Failure Snapshot Rural and Regional Centres • A total of 16 health services across the state participated in the 2016 Heart Failure Snapshot, with enrolment of 456 • The 2016 cohort comprised 148 patients with suspected patients. The majority (58%) were male, and the median STEMI, presenting to 9 rural or regional health services age was 76 years. Overall, 63% of patients had some form across Victoria. A total of 41 patients (28%) were ineligible of left ventricular dysfunction, with 46% having for thrombolysis. Four were triaged to primary PCI and predominantly systolic dysfunction and 14% predominantly transferred to a PCI capable hospital. The remaining 103 diastolic dysfunction. patients, all received thrombolysis, either at the treating hospital (n=94) or via a pre-hospital thrombolysis protocol, • The most common co-morbidities were atrial fibrillation administered by Ambulance Victoria (n=9). (54%), diabetes (47%), anaemia (36%), history of angina (34%), chronic obstructive pulmonary disease (COPD) • Only 57% of patients were transported by ambulance to /asthma (31%), and moderate chronic kidney disease hospital, while 36% were driven in by friends or family (self- (31%). presenters). Most (90%) patients were located within 50km of their treating hospital at symptom onset. • The majority of patients (85%) presented via the emergency department, with just under half (49%) • The median time from pain onset to first medical or admitted into a general medicine unit for management. ambulance contact was 120 minutes. The median time Only 3 hospitals had a dedicated heart failure unit. taken for an ambulance to arrive was 12 minutes and the Shortness of breath was the predominant reason for time to transfer to hospital was 52 minutes. admission (88% of cases).

The median time taken from hospital arrival to the • • For patients with heart failure with reduced ejection recording of the first ECG was 10 minutes – compliant with fraction, there were increases in the use of guideline- guideline recommendations. recommended medications at discharge.

The median door-to-needle time (time from patient arrival • • The median length of hospital stay was 6 days (IQR: 3, 9), to time thrombolytic drug administered) was 39 minutes with the majority of patients (72%) discharged to home. (IQR: 27, 69). None of the 9 participating hospitals managed to achieve a median door-to-needle time within the • The unadjusted in-hospital mortality rate was 6%, rising to Australian guidelines recommendation of ≤30 minutes, and 10% at 30 days follow-up. Mortality rates for all hospitals overall, only 32% of cases received timely thrombolysis were within control limits and there were no outliers in within 30 minutes. performance. Higher mortality rates were observed in patients with systolic dysfunction (9%), compared with • The in-hospital mortality (before transfer) for the overall patients with predominantly diastolic dysfunction (5%). cohort was 6.1%. Mortality was higher among patients with cardiogenic shock (13 patients, 7 deaths (54%). There were • At 30 days post discharge, the all-cause readmission rate 3 cases of major bleeding but no cases of stroke or intra- was 21% which was lower than in the 2015 Heart Failure cerebral haemorrhage. Snapshot (26%).

• Most thrombolytic-treated patients (95%) were subsequently transferred to a PCI capable hospital within 24 hours. The median time from referral request to the PCI capable hospital to the actual transfer from regional STEMI sites was 2.3 hours.

• Linkage with the VCOR PCI module determined that 85% of the entire cohort (both thrombolysis eligible and ineligible) were transferred to a PCI capable hospital. Of those, 68% A/Prof Jeffrey Lefkovits had a subsequent PCI and 9% underwent CABG surgery. VCOR Clinical Director

10 VCOR | 2016 Annual Report Introduction

The Victorian Cardiac Outcomes Registry (VCOR) is a clinical quality registry, established in 2012 to monitor the performance of health services in Victoria in their delivery of a range of cardiac-based therapies. VCOR first reported on percutaneous coronary intervention (PCI), and has subsequently established modules for early management of acute ST elevation myocardial infarction in non-metropolitan areas and acute treatment of heart failure in the hospital setting. The registry encompasses hospitals in both the public and private sector and reports on the quality and effectiveness of cardiovascular health care in Victoria.

As VCOR enters its fifth year of operation, the broad aim of VCOR is still to provide meaningful information to clinicians, hospitals, health funders and consumers. The registry was developed to facilitate health services in their own continuous service development. The data from the registry allows hospitals to review and benchmark their own performance, ensuring patients receive the highest quality cardiac care possible. More broadly, the registry can also be utilised to assess overall compliance with national standards of cardiac care and evidence-based guidelines as well as contribute to their continuing development and refinement.

The design and implementation of VCOR as a clinical quality registry is based around the Framework for Australian Clinical Quality Registries [3], developed by The Australian Commission for Safety and Quality in Health Care, in collaboration with the states and territories and expert registry groups. This framework was endorsed by the Australian Health Ministers’ Advisory Council (AHMAC) in March 2014. Its application provides assurance to all key stakeholders that registry data and its supporting systems satisfy minimum security, technical and operating standards.

Registry Governance and Structure

Governance Clinical Quality Committee

The governance of VCOR has been outlined in detail in The Clinical Quality Committee (CQC) has a key role in previous Annual Reports[4]. VCOR conforms to the National meeting the overall aim and purpose of VCOR. Quarterly Operating Principles for Clinical Quality Registries as set out review of hospital key performance indicators (KPIs) and by the Australian Commission on Safety and Quality in Health other data continued in 2016, with provision of numerous reports to sites across all modules. As in previous years, the Care. In 2016, a review of the Governance Terms of CQC was active in identifying and assisting to manage with Reference was undertaken and is awaiting ratification by the outlier performance, including providing feedback and Steering Committee. A review of committee membership review as requested. The robust processes related to the was also undertaken and ratified across all committees. management, review, benchmarking and reporting of data in VCOR is underscored by the work of the CQC. Steering Committee

The Steering Committee (SC) membership comprises Data, Research and Publications Committee representatives from each of the participating PCI sites, the Victorian Cardiac Clinical Network, the Department of The Data, Research and Publications Committee (DRP) Epidemiology & Preventive Medicine at Monash University reviews requests for analyses and access to group de- and a consumer representative. Additionally, the SC identified data for research projects. In 2016, the DRP expanded to include the Clinical Leads of the STEMI module reviewed several requests and approved two collaborative and the Heart Failure Snapshot module as members in 2016. research projects that required group de-identified data. The DRP also oversaw several abstracts based on VCOR data that will be outlined further in the next section of this report.

VCOR | 2016 Annual Report 11 Registry Activities

In 2016, the activities of VCOR as a clinical quality registry were directed to 3 specific areas in cardiovascular care. These were -:

1. Percutaneous coronary intervention

2. Early management of ST elevation myocardial infarction in rural and regional hospital settings

3. In-hospital management and outcomes of patients with acute decompensated heart failure

Each of these areas is dealt with in detail in subsequent sections of this report.

In addition to the regular quarterly reports issued to all participating hospital sites, VCOR completed its first Special Report to Sites, focussing on an area of particular clinical interest. For this report, VCOR had identified that hospitals were struggling to consistently achieve accepted benchmarks for time to treatment in patients with acute ST elevation myocardial infarction. The report’s intention was to highlight the issue, provide detailed and in-depth information on PCI hospital performance and provide tools to assist sites in improving their performance. The report was very well received and we look forward to monitoring this key process measure to see if the report positively influences performance. VCOR intends to continue this program of Special Reports into particular areas of interest to further assist hospitals in their continuous improvement programs.

As the day-to-day activities of data collection, analysis and reporting have become more streamlined and routine, VCOR has been able to increase its research activity related to safety and quality of care. With the support of Monash University’s Centre of Cardiovascular Research and Education in Therapeutics (CCRET) and collaboration with internal and external researchers, a number projects are underway or are being planned. These include the refinement and validation of risk-adjustment models for mortality, new models for patient-reported outcome measures, epidemiological aspects of cardiac disease and health economics. A list of research publications for 2016 relating to VCOR is included at the end of this Report.

Another new and exciting area of activity that has opened up is the development of data linkages with other datasets and data collection bodies. Data linkage, as a strategy, combines information about people across different databases to obtain a bigger and more complete clinical picture than would be possible from one database alone. It can increase the value and use of existing data holdings at relatively low cost and has enormous potential for clinical research. In 2016, data linkages were undertaken across VCOR modules to provide a more comprehensive picture of the patient journey. Collaborations were also established with Ambulance Victoria (AV), the Department of Health and Human Services (DHHS) and the and New Zealand Society for Cardiac and Thoracic Surgeons Database (ANZSCTS). These links will further define the patient journey across multiple service providers and episodes of care, and help to better understand longer-term patient outcomes.

12 VCOR | 2016 Annual Report Percutaneous Coronary Intervention (PCI)

Registry Module Activity

While this report primarily covers the 2016 calendar year, it also provides insights across the 4 years VCOR has been collecting data in selected areas of interest. PCI hospital participation is outlined in Table 1. In 2016, 25 PCI hospitals contributed data, including all 13 public hospitals and 12 of 17 private hospitals. The remaining five private hospitals were engaged in 2016 and will commence data collection in 2017, resulting in complete capture of all PCI data across Victoria.

Table 1: Participation of Victorian PCI hospitals

Victorian PCI hospitals Hospital type 2013 2014 2015 2016 Alfred Hospital, The Public • • • • Austin Hospital Public • • • • Public • • • • Bendigo Hospital Public • • • • Public • • • • Cabrini Hospital Malvern Private • • • • Epworth Hospital Eastern Private • • • Epworth Hospital Richmond Private • • • • Epworth Hospital Geelong Private • Public • • • • Geelong Private Hospital Private • • • Jessie McPherson Private Hospital Private • • • • Linacre Private Hospital Private ° Knox Private Hospital Private • • • • Melbourne Private Hospital Private • • • MonashHeart, Public • • • • Peninsula Private Hospital Private ° St John of God Hospital (Ballarat) Private ° ° St John of God Hospital (Bendigo) Private • • St John of God Hospital (Geelong) Private ° ° St Vincent's Hospital Melbourne Public • • • • St Vincent's Private Hospital Private • • • • The Northern Hospital Public • • • • The Public • • • • The University Hospital, Geelong Public • • • • The Valley Private Hospital Private • Warringal Private Hospital Private ° Western Hospital (Footscray) Public • • • • Western Hospital (Sunshine) Public N/A N/A • • Western Private Hospital Private • • • • Table Legend: • = contributing data; ° = engaged but not yet contributing

A total of 9,992 cases were entered into VCOR in 2016, involving 8,893 patients, with 11% of patients undergoing more than one procedure. The number of cases treated in the public sector was 6,216 (62%) and 3,776 in private (38%). All cases included in this report had completed baseline and follow-up data available for reporting. The lost-to-follow-up (LTF) rate for 2016 was 1.5%, and the overall rate for the entire registry since its commencement is 1%. At the end of 2016, the VCOR registry had accumulated 32,353 cases, as shown in Figure 1.

VCOR | 2016 Annual Report 13 Figure 1: Cumulative cases submitted by month from 2013-2016

35, 000 Pr ivate Pu blic 30, 000

) 25, 000 N (

s r e

b 20, 000 m u n

e 15, 000 s a c

I

C 10, 000 P

5, 000

0 3 3 3 4 6 4 5 6 4 5 6 5 3 3 3 4 6 4 5 6 3 4 6 5 5 4 6 5 3 4 3 6 6 5 3 5 4 3 6 5 4 4 6 5 3 6 4 5 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ------l l l l t t t t r r r r r r r r c c c c y y y y v v v v g g g g b p b p b p p b n n n n n n n n u u u u c c c c a a a a p p p p e e e e a a a a o o o o u u u u a a a a e e e e e e e e u u u u J J J J J J J J J J J J O O O O A A A A F S F S S F S F D D D D A A A A N N N N M M M M M M M M

Data Quality - Audit Activity

A key operational activity of a clinical quality registry is the performance of regular audits to ensure accuracy and completeness of data collection and entry into the registry[5]. This generally encompasses assessment of eligible cases to ensure all relevant cases are entered, as well as verification of source data. VCOR commenced auditing PCI cases in March 2014 and the process is ongoing. Case ascertainment audits (assessing that all eligible cases are entered into the registry) are undertaken on a yearly basis. Most sites have now undergone multiple audits since their entry into the registry. Figure 2 below illustrates progressive improvement in hospital performance with successive audits.

In instances where missing cases were identified, sites were subsequently able to retrieve and enter these cases into the registry. Only two sites had missing cases greater than 5% and these sites have subsequently instituted process changes to improve case entry compliance.

Data Quality assessment (review of source data to Figure 2: Overall rate of missing cases across each annual determine accuracy), are undertaken every 3 years whereby 5% of case records are randomly selected audit (2014-2016) for comparison with the hospital medical record. 4 Two additional brief audits were undertaken in 2016. The overall agreement rate between VCOR data and the hospital medical record is 97.5% from audits at 3 ) %

23 sites, indicating high quality data collection. This (

g n compares favourably with national and international i s s registries [6, 7]. i 2 m

s e s a

All sites are given a detailed, individualised report of C their audit findings. VCOR recognises that registry 1 data must be high quality and is committed to an ongoing, thorough and accurate audit program. 0 Audi t 1 Aud it 2 Aud it 3

14 VCOR | 2016 Annual Report Patient Characteristics

For the 2016 cohort, the median age for males was 65 years (IQR: 57, 74) and for females 70 years (IQR: 61, 78). The distribution of cases by age is similar to previous years. As shown in Figure 3, the peak frequency of cases among men is in the seventh decade and for women, it is in the eighth decade. Table 2 compares selected patient demographic information from 2016 with previous years and indicates that the demographic profiles of the patients have remained similar across the four-year period.

Figure 3: Age and gender distribution of patients undergoing PCI

35

30

25

) 20 % (

s e

s 15 a C 10

5

0 ≤4 0 yrs 41-50 yrs 51-60 yrs 61-70 yrs 71-80 yrs 81-90 >9 0 yrs Ag e Group Ma le (n=765 1) Fe male (n=2341)

Table 2: Selected patient characteristics 2013-2016

Patient characteristics 2013 2014 2015 2016 (N=4809) (N=8327) (N=9225) (N=9992) Age - years (Mean ±SD) 65.8 (±12.0) 65.3 (±11.9) 65.6 (±12.0) 66.0 (±12.0) % % % % Gender - female 22.9 23.1 23.0 23.4 Diabetes 22.4 21.6 23.0 21.7 PVD History 3.5 3.8 3.6 3.5 CVD History 3.8 3.7 4.0 3.3 Previous PCI 34.8 31.9 32.9 32.7 Previous CABG 8.9 8.4 7.6 7.6

Patient demographic factors by hospital sector are shown in Table 3. Patients treated in private hospitals tended to be older and were more likely to have had a previous PCI. The difference in age profiles between patients in the public and private sectors is illustrated in Figure 4. While elderly patients (>80 years) accounted for 11.8% of the overall PCI cohort, the proportion of elderly patients in the private sector was almost double the public rate (16.6% vs 8.9%). As in previous years, the elderly had higher rates of vascular comorbidities (peripheral vascular disease, cerebrovascular disease and previous bypass surgery), with an approximate doubling of these incidences compared with patients under 80 years (PVD: 6.8% vs 3.0%, CVD: 6.3% vs 2.9% and CABG: 12.7% vs 6.9%).

VCOR | 2016 Annual Report 15 Table 3: Selected patient characteristics by hospital sector

Patient characteristics Public Private (n=6216) (n=3776) Age - years (Mean ±SD) 63.8 (±12.1) 69.6 (±10.9) % % Gender - female 22.5 24.9 Diabetes 22.6 20.2 PVD History 3.3 3.8 CVD History 3.5 2.9 Previous PCI 26.7 42.6 Previous CABG 6.4 9.4

Figure 4: Age distribution for public and private patients

35

30

25

20 ) % (

s

e 15 s a C 10

5

0 ≤4 0 yrs 41-50 yrs 51-60 yrs 61-70 yrs 71-80 yrs 81-90 >9 0 yrs Ag e Gro up Pu blic (N=6216) Pr ivate (n =3776)

Clinical Presentation Figure 5: Procedures by clinical presentation

Just over half the patients in the 2016 PCI cohort presented with an acute coronary syndrome (ACS). This includes the conditions of ST elevation myocardial infarction (STEMI), non-ST elevation STEMI myocardial infarction and unstable angina. The latter two are often 22% grouped together as non-ST elevation ACS (NSTE-ACS). The percentages of patients with either an ACS or stable coronary Non-ACS disease across the VCOR cohort are shown in Figure 5. 47% NSTE-ACS 31%

16 VCOR | 2016 Annual Report While the proportion of ACS presentations has remained fairly constant over the previous 4 years, there are distinct differences among hospitals in regard to the burden of ACS in their overall case mix (Figure 6). These differences are highlighted when clinical presentation is categorised by hospital type. As shown in Figure 7, the public sector treats almost double the proportion of ACS patients compared with the private sector. This difference arises because the vast majority of PCIs for STEMI are managed in public hospitals in Victoria. This trend has been observed in previous years as well.

Figure 6: ACS and non-ACS cases by hospital

100 90 80 70 60

) %

( 50

s e

s 40 a C 30 20 10 0 AB CDEFGH I J KLMNOPQRSTUVWXY Ho spital AC S No n-ACS

Figure 7: Procedures by clinical presentation for public and private hospitals

STEMI 8%

STEMI Non-ACS 30% 34% NSTE-ACS 25% Non-ACS 67%

NSTE-ACS 36%

Public Private

VCOR | 2016 Annual Report 17 Indications for PCI

There is a growing body of evidence that suggests/supports monitoring reasons for performing PCI cases positively influences the overall quality of health care delivery. American peak cardiology organisations in particular, have developed guidelines for the appropriate use of PCI in both stable coronary disease [8] and acute coronary syndromes[9]. Yet, these appropriateness use criteria utilise comprehensive sets of criteria that are quite complex and their overall system has a number of other potential limitations to its implementation here in Victoria. As VCOR was designed around a utilitarian model focused on the practical assessment of procedural and patient outcomes, the registry has taken a simpler approach of providing a straightforward breakdown analysis of the indications for PCI to further understand and evaluate the reasons for performing PCI.

For ACS patients, revascularisation recommendations are generally based on the large body of evidence that supports an invasive revascularisation strategy. A diagnosis of ACS is therefore considered a primary indication for PCI in suitable patients with an identifiable culprit lesion, in accordance with Australian practice guidelines [10]. Table 4 outlines the proportions of various categories and sub-categories of ACS as the indication for PCI. Primary PCI for acute STEMI accounted for 28% of all ACS indications with the vast majority of these performed in the public sector. The largest ACS indication for PCI overall was NSTEMI (42%), with three-quarters of the caseload treated in public sector. The proportions of patients in the various sub-categories of ACS are similar to previous years.

Table 4: PCI Indication by ACS category and sub-category

A l l s i t e s ( N = 5 7 0 3 ) P u b l i c ( n = 4 2 3 9 ) P r i v a t e ( n = 1A4ll6 s4i)tes Public Private ACS category N ( % ) N ( % ) N ( % ) (N=5703) (n=4239) (n=1464) ACS category N (%) N (%) N (%) Primary PCI for acute STEMI 1603 (28.1) 1374 (32.4) 229 (15.6) STEMI PCI 12-24 hours after symptom onset 230 (4.0) 203 (4.8) 27 (1.8) Pharmaco-invasive PCI 228 (4.0) 200 (4.7) 28 (1.9) Rescue PCI 99 (1.7) 91 (2.1) 8 (0.5) PCI for OHCA/shock (non-MI) 31 (0.5) 28 (0.7) 3 (0.2) PCI for NSTE-ACS 3512 (61.6) 2343 (55.3) 1169 (79.8) NSTE-ACS sub-category N (%) N (%) N (%) NSTEMI 2394 (42.0) 1793 (42.3) 601 (41.1) UAP 748 (13.1) 408 (9.6) 340 (23.2) Recent ACS 7-30 days ago 370 (6.5) 142 (3.3) 228 (15.6)

When assessing appropriateness in stable (non-ACS) coronary disease, there is greater emphasis on the use of objective measures of ischaemia and stratification of patients into low, medium or high-risk groupings through lesion severity and presence of symptoms. Table 5 provides a breakdown of the reasons for PCI among non-ACS patients. About two-thirds of non-ACS patients were treated for stable angina, with a slightly smaller proportion in the private (63%) compared with the public (70%). A significant proportion (20%) of cases were attributed to the indication of “staged” PCI. This group of patients, with multi-vessel disease undergoing a second (and sometimes a third or fourth) PCI, was heterogeneous, with some patients with a recent ACS, others with an ACS more than 30 days ago, and others with no recent history of ACS. In these cases, it was more difficult to ascribe the usual indicators for PCI including angina symptoms, positive functional test and high-grade coronary stenosis (>70%).

Table 5: PCI Indications for non-ACS cases

All sites Public Private (N=4289) (n=1977) (n=2312) N (%) N (%) N (%) Stable angina 2845 (66.3) 1392 (70.4) 1453 (62.8) No symptoms and no functional test 230 (5.4) 96 (4.9) 134 (5.8) No symptoms and positive functional test 259 (6.0) 84 (4.2) 175 (7.6) Staged PCI after ACS (<30 days after first procedure) 399 (9.3) 230 (11.6) 169 (7.3) Staged PCI after ACS (>30 days after first procedure) 136 (3.2) 87 (4.4) 49 (2.1) Staged PCI after original non-ACS indication 335 (7.8) 65 (3.3) 270 (11.7) Miscellaneous* 85 (2.0) 23 (1.2) 62 (2.7)

*Miscellaneous group include arrhythmia, cardiomyopathy/heart failure, syncope, second attempt at lesion and others

18 VCOR | 2016 Annual Report Cases were examined to determine how many of the three clinical, functional and anatomic indicators were present. Symptoms of angina or its equivalent were reported in 83% of non-ACS patients. A high-grade stenosis was noted in 92% and a positive functional test in 42%. A total of 87% of non-ACS patients had at least 2 of 3 indicators (Table 6).

Table 6: Non-ACS patients: Clinical Indicators for PCI

S y m p t o m s P o s i ti v e f u n c ti o n a l t e s t H i g h g r a d e s t e n o s i s T o tal N (%) • • • 1069 (31.3) ° • • 257 (7.5) • • ° 84 (2.5) • ° • 1559 (45.6) • ° ° 133 (3.9) ° ° • 267 (7.8) ° • ° 18 (0.5) ° ° ° 32 (0.9) 3419 (100) )Table Legend: • = clinical indicator present; ° = clinical indicator not present Excludes all staged PCI cases (n=870)

There was some variation in the pattern of PCI indicators when the public and private sectors were compared. Figure 8 demonstrates that there was a weighting towards 2 clinical indicators in the private (60% of non-ACS cases) compared with the public sector (50% of non-ACS cases) .

Figure 8: Clinical indicators for PCI in non-ACS patients for public and private hospitals

No Clinical Indicator No Clinical Indicator 1 Clinical Indicator 1% 1 Clinical Indicator 1% 9% 15%

Public Private

VCOR | 2016 Annual Report 19 Clinical and Lesion Subsets

Trends in PCI management over time are reflected by the incidence and treatment of specific types of coronary lesions. In particular, there has been a very strong worldwide trend in the last decade to undertake PCI for lesions in the left main coronary artery, previously almost exclusively the domain of cardiac surgery[11]. Among the VCOR cohort, left main PCI still remains an infrequent procedure, accounting for only 2% of cases. PCI for other lesion subsets including multi-vessel disease (7.2%) and in- stent restenosis (5.3%) was also relatively uncommon, with rates similar to the last four years (Figure 9). Advanced techniques for treatment of chronic total occlusions is another area of burgeoning development, but case numbers have not really increased yet, accounting for just 4% of the overall workload. However, this is expected to grow in the coming years and VCOR will be focussing on this area through the monitoring of uptake, performance and outcomes with these highly specialised PCI techniques.

Figure 9: Comparative trends in incidence over time for selected PCI lesion subsets

8 20 13 7 20 14 20 15 6 20 16 5 ) % (

4 s e s a

C 3

2

1

0 Le Ma in Mu lvessel In-s tent Ch ronic T ot al Co ronar y A rtery PC I Re stenosis Oc cl usion Le sio n su bse t

Another trend in PCI management has been an increasing interest in high-risk PCI for cardiogenic shock. In 2016, the proportion of cases presenting with cardiogenic shock or out-of-hospital cardiac arrest was 3.1% and has not substantially increased in the last four years (Table 7).

Table 7: Patients presenting with cardiogenic shock or out-of-hospital cardiac arrest (OHCA): 2013-2016

Presentation type 2013 2014 2015 2016 (N=4809) (N=8327) (N=9225) (N=9992) N (%) N (%) N (%) N (%) Cardiogenic shock 82 (1.7) 198 (2.4) 223 (2.4) 278 (2.8) Intubated OHCA 36 (0.7) 89 (1.1) 108 (1.2) 126 (1.3) Shock and/or intubated OHCA 95 (2.0) 239 (2.9) 253 (2.7) 309 (3.1)

20 VCOR | 2016 Annual Report Device Use

VCOR has also followed patterns of device use in PCI that can evolve over time. In Figure 10: Device use among PCI cases 2016, at least one coronary stent was implanted in 94% of cases, slightly more than the previous year (92%). Balloon angioplasty alone (POBA) was performed in a very Any BVS small number of cases, as was drug-eluting balloon angioplasty to treat in-stent POBA Other 1% restenosis. Despite the availability of the bioresorbable vascular scaffold (BVS) in 3% 3% the Australian market since 2013, the uptake of this new stent technology has been small. In 2016, just 72 patients (1%) across 15 hospitals received a BVS, similar to BMS the previous year. A breakdown of device use in 2016 is shown Figure 10. 8%

The total number of stents used and total stent length per patient are shown in Figures 11 and 12. The majority of patients (67%) received a single stent and in the majority of cases (66%), the total length of stents that were deployed was between DES 11-30mm. Multiple or overlapping stents totalling >50 mm in length were used in 85% 8% of cases (Figure 12), equally distributed across the public and private sectors.

Figure 11: Number of stents used per PCI case (includes all lesions treated)

70

60

50

40

30

C 20

10

0 01 234 >4 Num be r of stent s used pe r PCI

Figure 12: Total length of stents used per PCI case (includes all lesions treated)

25

20

15

10 C

5

0 01 -10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65 66-70 >70

To tal length of stents per PCI (mm)

VCOR | 2016 Annual Report 21 The use of various adjunctive devices in PCI was generally limited (Table 8). Devices that assist the visualisation of the coronary anatomy such as intravascular ultrasound (IVUS) and optical coherence tomography (OCT) were employed infrequently, but were more commonly used in the public sector. On the other hand, fractional flow reserve (FFR) - an evidence-based technique used to determine whether coronary lesions should be treated - was applied equally across the public and private systems. Rotational atherectomy - a mature adjunctive treatment particularly useful for heavily calcified lesions - was used sparingly in 2016, as in previous years.

Recent data have discounted the effectiveness of routine thrombus aspiration (catheter-based extraction of blood clots that occlude the coronary artery) in the setting of acute STEMI[12]. In 2016, thrombus aspiration occurred in 3.5% of cases – mostly in STEMI cases (n=289), but also in a small number of other cases including NSTE-ACS (n=45) and non-ACS (n=12). This technique was utilised mostly in the public sector. VCOR also tracked the use of intra-aortic balloon pump (IABP) and extracorporeal membrane oxygenation (ECMO) devices that support the circulation in patients with cardiogenic shock. The rate of use of these devices was low and they were almost exclusively deployed in the public hospital system.

Table 8: Adjunctive device use

Adjunctive device type All sites(N=9992) Public (n=6216) Private(n=3776) All sites Public Private Adjunctive device type (N=9992) (n=6216) (n=3776) N (%) N (%) N (%) Intravascular ultrasound 118 (1.2) 92 (1.5) 26 (0.7) Optical coherence tomography 60 (0.6) 51 (0.8) 9 (0.2) Thrombus aspiration device 346 (3.5) 298 (4.8) 48 (1.3) Rotational atherectomy 101 (1.0) 59 (0.9) 42 (1.1) Fractional flow reserve 254 (2.5) 158 (2.5) 96 (2.5) IABP 71 (0.7) 64 (1.0) 7 (0.2) ECMO 12 (0.1) 10 (0.2) 2 (<0.1)

Drug Eluting Stents

Drug-eluting stents (DES) are now the predominant stent type used in all hospitals across Victoria (Figure 13). They have demonstrated superiority to bare metal stents in relation to the risk of stent re-narrowing (restenosis), and previous concerns about their safety have largely been dispelled, especially with the second-generation drug-eluting stents now commonly used. The registry has documented continuing growth in their use from around 75% in the previous three years up to 85% in 2016. Nevertheless, there is still significant variation in DES use among hospitals, ranging from 59% to 96% as shown in Figure 13. The use of DES has been greater in the private sector, although the gap is narrowing (Figure 14).

Figure 13: DES use by hospital

100 90 80 70 60 ) % (

50 s e s

a 40 C 30 20 10 0

Hosp ital

22 VCOR | 2016 Annual Report Figure 14: Trends in DES usage by hospital sector: 2013-2016

100 90 80 70 60 ) % ( 50 s e s

a 40 C 30 20 10 0 2013 2014 2015 2016 Ye ar Pu blic Pr ivate

The use of DES in acute STEMI (79%) still remains lower than in other acute coronary syndromes (85%) and in non-ACS stable disease (88%) (Figure 15). However, the rate in acute STEMI has increased compared with the previous year (64% in 2015). Smaller increases in DES use were observed in the NSTE-ACS cohort (78% in 2015) and the non-ACS cohort (82% in 2015).

Figure 15: Device use by clinical presentation

100 90 9 4 15 80 70 60

) 50 % (

s 85 88 e 40 79 s a C 30 20 10 0 ST EMI NS TE-ACS No n- ACS (n =2171) (n =3141) (n =4680) Clin ic al Presentao n An y DES BM S PO BA An y BVS Ot her

VCOR | 2016 Annual Report 23 Arterial Access

The use of the radial artery for arterial access continued to grow in VCOR hospitals in 2016. Previous VCOR reports demonstrated that patterns of practice in arterial access have been changing. In 2016, for the first time, radial access was more commonly used than the femoral route (55% radial vs 45% femoral, Figure 16). This represented an absolute 10% increase over the previous year, and is reflective of the worldwide trend towards a predominance of radial access for PCI.

Figure 16: Trends in arterial access: 2013-2016 80

70

60

50 ) %

( 40

s e s

a 30 C 20

10 0. 1% 0. 3% 0. 1% 0. 1% 0 20 13 20 14 20 15 20 16 Ye ar Ra dial Fe moral Br achi al

Despite the overall increase in radial artery utilisation, rates of radial access varied widely among health services, ranging from 11% to 78% (Figure 17). In particular, the use of the radial artery was lower among private hospital patients (Figure 18). Yet, with the overall increase in radial artery use, both the public and private increased by similar percentages (21% and 22% relative increases) over the previous year.

Figure 17: Arterial access route by hospital

100 90 80 70 60 ) % (

50 s e s

a 40 C 30 20 10 0 AB CDEFGH I J KLMNOPQRSTUVWXY Ho spital Ra dial Fe moral Br achi al

24 VCOR | 2016 Annual Report Females tended to have the radial artery used less often than males (48% vs 57%), but the overall percentage in females had increased over the previous year (40% in 2015). In the elderly, lower rates of radial access were seen (43% >80 years vs 57% <80 years).

Figure 18: Arterial access route in public and private hospitals

Brachial Brachial <1% <1%

Femoral Femoral Radial 37% Radial 63% 57% 43%

Public Private PCI for Acute STEMI

More than any other indication for PCI, acute STEMI places significant demands on the systems of care to treat patients in an effective, time critical manner. Hospitals often have to juggle limited resources to provide around-the-clock emergency service. Yet, they are mandated to comply with state and national performance standards in the delivery of timely and efficient emergency treatment. Accordingly, PCI for acute STEMI represents a special case for analysis and performance assessment by a clinical quality registry.

In 2016, 1,603 patients presented directly to a PCI hospital with an acute STEMI and underwent PCI as the primary reperfusion treatment for their heart attack. This represents 16% of the total caseload across the registry for the year. Cases were predominantly treated in the public hospital sector (86% of all STEMI cases), and they accounted for 26% of the public hospital workload. In contrast, PCI of STEMI accounted for 7% of the workload in the private sector. These percentages are similar to previous years. A breakdown of the proportion of cases treated for acute STEMI across hospitals is shown in Figure 19. A number of hospitals do not offer acute PCI services and therefore have zero cases.

Figure 19: Acute STEMI cases as a proportion of overall case numbers by hospital

45

40

35

) 30 % (

s

e 25 s a C

I 20 M E T

S 15

e t u

c 10 A 5 0% 0% 0 AB CDEFGH I J KLMNOPQRSTUVWXY All Hos pital

VCOR | 2016 Annual Report 25 Compared with the rest of the cohort, acute STEMI patients were younger (62.5 ± 12.6 years vs 66.8 ± 11.7 years), had fewer traditional cardiac risk factors such as diabetes (15.6% vs 23.1%), peripheral vascular disease (2.2% vs 3.7%) and cerebrovascular disease (2.7% vs 3.4%). The acute STEMI cohort also had fewer previous revascularisation procedures including previous PCI (11.9% vs 37.4%) and coronary artery bypass grafting (1.7% vs 8.9%). Privately treated patients were an older group than public sector patients and had a greater proportion of women (26% private vs 20% public).

Figure 20 shows the proportions of PCI treatment type for acute STEMI by hospital. The proportion of acute STEMI cases treated with a pharmaco-invasive approach for the entire cohort rose to 8.9%, almost 3% higher than the previous year. This strategy mainly applies to patients who initially present to a non-PCI hospital and undergo emergency reperfusion therapy with a clot-dissolving thrombolytic drug. They are then transferred urgently to a PCI hospital to undergo adjunctive PCI where suitable within 4-24 hours of the heart attack. The growth in this combination approach likely reflects recent Australian guidelines that encourage non-PCI hospitals to transfer their STEMI patients early, rather than continue to manage them in hospitals without on-site angiography facilities[13]. There has also been a significant change in transfer policy by Ambulance Victoria to facilitate early and rapid transport of STEMI patients.

Figure 20: PCI treatment type for acute STEMI patients by hospital

100 90 80

) 70 % (

s

e 60 s a c

I 50 M E

T 40 S

e t

u 30 c A 20 10 0 AC DEFGHI J KLMNOPQRSTVWXY Ho spital Pr imary PC I for STEMI Ph armaco-invasive Re scue PC I Low rates of acute STEMI cases for hospitals J, N, Q, Y (n<5), R and W (n<15)

The proportion of acute STEMI cases treated via radial arterial access increased by 13% since 2015, to 60% of STEMI cases in 2016. Given the strong evidence base for better outcomes when radial access is utilised in acute STEMI[14], the growth in radial artery usage in this patient cohort is appropriate. However, there is still some notable variation in radial use in the STEMI cohort across hospitals (Figure 21).

Figure 21: Radial access rates in acute STEMI cohort 100 90 80 )

% 70 (

s e s 60 a C

I

M 50 E T S 40 e t u c 30 A 20 10 0% 0% 0% 0 AC DEFGHI J KLMNOPQRSUVWXY All

Ho spital

Low rates of acute STEMI cases for hospitals J, N, Q, Y (n<5), R and W (n<15) 26 VCOR | 2016 Annual Report Door-to-balloon Times for PCI for Acute STEMI (Primary PCI)

The time taken from a STEMI patient’s arrival at the hospital till the insertion of a device to unblock the vessel (typically a balloon catheter) is known as the door-to-balloon time (DBT). It is a key performance indicator and provides an aggregate measure of how well hospitals’ systems and processes for managing time-critical STEMIs are working. When benchmarking hospitals, VCOR has adopted the national benchmark standard of a DBT ≤90 minutes[13], similar to other international guidelines[15]. This analysis is limited to STEMI cases presenting to a PCI capable hospital, and excludes STEMI patients transferred from outside hospitals and those who developed a STEMI while an in-patient.

The median door-to-balloon time for the overall primary PCI cohort was 67 minutes (IQR: 47, 96) (Table 9). The range for median door-to- balloon-times was 48-104 minutes, with all but one hospital achieving a median door-to-balloon time within the recommended ≤90 minutes benchmark (Figure 22).

Table 9: Door-to-balloon-time for primary PCI cases

Door-to-balloon Time Primary PCI * (all cases) Primary PCI * Primary PCI* Primary PCI* Primary PCI* Door-to-balloon Time (all cases) (PHN only ϯ) (no-PHN ϯ) (N=1303) (n=717) (n=586) Median – mins (IQR) 67 (47, 96) 55 (40, 76) 87 (61, 114) Proportion of cases ≤90 mins (%) 71.9% 86.3% 54.3%

*Primary PCI for STEMI presentations excluding all inter-hospital transfer arrivals ϯPre-hospital notification (PHN)

Figure 22: Door-to-balloon time for primary PCI cases by hospital

0 Target D BT ( 90 m ins) 0 2 All s ite s m edian D BT 0 ) 5 s 1 n i m (

e m i t

n o o l l 0 a 0 b 1 - o t - r o o d

n a i d e 0 M 5 0

AC DEFGH I KLMOPSTVWX Ho spital

Hospitals B, J, N, Q, R, U, Y not included (Primary PCI cases n<5) Hospital W low numbers (n<10)

VCOR | 2016 Annual Report 27 The median door-to-balloon (DBT) time, while useful as a benchmark measure, tends to over-estimate the performance of health services. Arguably, the compliance benchmark of ≥75% cases achieving DBT within 90 minutes is the preferred measure of performance in this context. The DBT compliance for the overall cohort was 72%, close to the international benchmark target of ≥75% of cases. However, there was a significant spread among hospitals ranging from 31% to 88% (Figure 23). While only 9 hospitals accomplished the compliance benchmark of ≥75% of cases, this was an improvement over 2015, when only 4 hospitals achieved the benchmark level.

Figure 23: Proportion of primary PCI cases with door-to-balloon time ≤90 minutes by hospital 100 Target: 75% ≤90 mins 90

) 80

%

(

s

n 70

i

m

0 60

9

n

i 50

h

t

i w

40

d

e t

a 30

e

r

t

s 20

e

s a

C 10 0 A C D E F G H I K L M O P S T V W X Total Hospital

Hospitals B, J, N, Q, R, U, Y not included (Primary PCI cases n<5) Hospital W low numbers (n<10)

Factors Influencing Door-to-Balloon Time Compliance Pre-hospital notification

Notification by ambulance to hospitals of an acute STEMI patient’s imminent arrival is referred to as pre-hospital notification (PHN). This allows hospitals to prepare for the arrival of a STEMI patient, activate the cardiac catheterisation laboratory team and set up rapid transfer from hospital entrance to catheter laboratory to minimise delays to the commencement of the PCI.

Figure 24 compares the proportion of primary PCI cases achieving DBT within 90 minutes for patients triaged with and without pre-hospital notification for each hospital. Compliance rates for DBT≥90 mins were higher in patients triaged with pre-hospital notification, with 86% achieving a door-to-balloon time of less than 90 minutes. In contrast, only 54% achieved DBT within 90 minutes without pre-hospital notification. There was significant variation in the number of cases with PHN across hospitals (range 15% - 83%) with an average of 55% for the overall cohort.

Figure 24: Proportion of primary PCI cases with door-to-balloon time ≤90 minutes - pre-hospital notification (PHN) vs no pre-hospital notification 100 90

) 80 % (

s

n 70 i m

0 60 9

n i

h 50 t i w 40 d e t

a 30 e r t

s

e 20 s a C 10 0 AC DEFGHI KLMOPSTVWX All Ho spital PH N No PHN Hospitals B, J, N, Q, R, U, Y not included (Primary PCI cases n<5) Hospital W low numbers (n<10) 28 VCOR | 2016 Annual Report In-hours versus out-of-hours presentation The proportion of STEMI cases treated out-of-hours and therefore necessitating staff call-backs was 61% for the overall cohort. As in previous reports, cases performed outside normal working hours demonstrate longer delays till commencement of the procedure (Figure 25). This was predominantly due to increases in the time taken to transfer patients to the cardiac catheter laboratory. The proportion of hospitals’ STEMI workload that fell out-of-hours ranged from 21%-90%.

Figure 25: Proportion of primary PCI cases with door-to-balloon time ≤90 minutes - in-hours vs out-of-hours presentation 100 90

) 80 % (

s 70 n i m

0 60 9

n i 50 h t i w

40 d e t

a 30 e r t

s 20 e s a

C 10 0 AC DEFGHI KLMOPSTVWX All Ho spital In -h ours Ou t-of-hours Hospitals B, J, N, Q, R, U, Y not included (Primary PCI cases n<5) Hospital W low numbers (n<10) In-hours: 8:00am – 6:00pm (Mon – Fri). Out-of-hours: 6:00pm – 8:00am (Mon – Fri and weekends).

Outcomes

Lesion and Procedure Success Rates In general, a PCI is considered successful if the narrowed or blocked coronary arteries are re-opened with a coronary stent or balloon and the patient remains free of complications. Lesion success is defined as a reduction of the narrowing to less than a 10% residual stenosis following stenting, or less than 50% following balloon angioplasty alone. Procedure success is defined as all attempted lesions successfully treated, without any significant in-hospital complications. As in previous years, the lesion success rate was high at 95% and the overall procedure success rate was 91% (range for hospitals 83%-99%).

New Renal Impairment The use of angiographic contrast agents during PCI can lead to acute impairment of renal function, especially in patients with pre- existing kidney disease, diabetes, hypertension or advanced age. Steps can be taken to minimise the risk of contrast-induced nephropathy and its incidence is an important outcome measure of any PCI service.

The overall rate of new renal impairment, defined as a serum creatinine rise >44.2 μmol/L or 25% above pre-procedural value within 5 days of PCI procedure was 3.3%. An analysis of new renal impairment rates for selected high-risk subgroups is shown in Figure 26. For each of the high-risk subgroups, the rate of new renal impairment was between 1.5-10 times higher than its comparator.

New renal impairment (RI) also varied according to clinical presentation. The highest incidence of new renal impairment was in STEMI cases (6.2%), which often involve patients with the greatest acuity and periods of hypotension (low blood pressure) that exacerbate the risk of renal impairment (Figure 27).

VCOR | 2016 Annual Report 29 Figure 26: Rate of new renal impairment in selected high-risk subgroups

40

35

30

25 ) % (

20 s e s

a 15 C

10

5

0 Ag e (>80yrs) Pre-pr oc RI Diabetes Shoc k/OHCA IA BP/ECMO Su bgr oup Ri sk factor pre sent Ri sk factor not present

Data available for 7165 cases

Figure 27: New renal impairment rates by clinical presentation

7

6

5

4 ) % (

s

e 3 s a C 2

1

0 ST EMI NST E-A CS non -ACS Al l c as es AC S Group Data available for 7165 cases

Referrals to Cardiac Rehabilitation

National guidelines and clinical care standards strongly endorse referral of patients with coronary disease to an appropriate cardiac rehabilitation or other secondary prevention program. Yet, less than 50% of patients with an ACS were referred to cardiac rehabilitation on a recent national survey and compliance with secondary prevention measures was poor[16]. VCOR has therefore monitored the rate of referrals after PCI to assist hospitals in their discharge planning and secondary prevention strategies.

In 2016, the overall rate for referral to cardiac rehabilitation after an admission for PCI among VCOR hospitals was 76.3%. Table 10 demonstrates that referral rates varied according to clinical presentation – highest for patients with STEMI and lowest in stable non- ACS presentations. Referral rates also varied by hospital sector – 84% public sector vs 64% private sector. There are likely several factors affecting these referral rates and is an area that will require further attention and analysis.

30 VCOR | 2016 Annual Report Table 10: Referral rates to cardiac rehabilitation by clinical presentation

Cases with Rehabilitation Clinical Presentation data available referral rate N N (%) STEMI 2034 1740 (85.5) NSTE-ACS 3112 2457 (78.9) Non-ACS 4654 3283 (70.5) All cases 9800 7480 (76.3)

Compliance with Guideline-Recommended Discharge Medications

Dual anti-platelet therapy and statins are recommended therapies for all patients undergoing PCI and stent insertion unless there are specific contraindications. Compliance with these medications was high, with 94% of patients discharged on dual anti-platelet therapy and a statin. For patients presenting with an ACS that have additional drug requirements, Table 11 and 12 outlines the rates of prescription of these guideline-recommended medications after PCI. The highest use was in STEMI patients. When considering the 5 main evidence-based drugs recommended after a PCI associated with an ACS (aspirin, additional anti-platelet, statin, beta blocker and ACE inhibitor/ARB), the proportion of patients discharged on at least 4 out of 5 medications was 83%.

Table 11: Discharge Prescription Rates of Beta Blockers (BB) by clinical presentation

Cases with Clinical Presentation data available BB rate N N (%) STEMI 2014 1776 (88.2) NSTE-ACS 3067 2301 (75.0) Non-ACS 4570 2731 (59.8) All cases 9651 6808 (70.5)

Table 12: Discharge Prescription Rates of ACE Inhibitor/Angiotensin Receptor Blockers (ARB) by clinical presentation

Cases with Clinical Presentation data available ACEI/ARB rate N N (%) STEMI 2009 1690 (84.1) NSTE-ACS 3080 2242 (72.8) Non-ACS 4569 2985 (65.3) All cases 9658 6917 (71.6)

VCOR | 2016 Annual Report 31 Key Performance Indicators

VCOR reports on a number of key performance indicators (KPIs) based on clinically relevant procedural outcomes. The KPIs reported for the VCOR PCI module are:

• In-hospital mortality • In-hospital major bleeding • Length of stay • In-hospital unplanned revascularisation • Door to balloon/device time for STEMI patients • 30-day risk-adjusted mortality • 30-day major adverse cardiac and cerebrovascular event (MACCE)

For 30-day risk adjusted mortality rates a risk-prediction scoring tool, based on the Melbourne Interventional Group[17] multicentre PCI registry mortality risk-adjustment model was developed and has been used by VCOR since 2014. The clinical characteristics used to construct the VCOR risk-adjustment model were:

• Age ≥80 years • Acute coronary syndrome • Glomerular filtration rate • Left ventricular ejection fraction • Cardiogenic shock • Left anterior descending coronary artery disease

In-hospital Mortality The unadjusted in-hospital mortality rate for 2016 was 1.8%. Table 13 demonstrates that the highest mortality was found in patients presenting with cardiogenic shock or out-of-hospital cardiac arrest requiring endotracheal intubation (OHCA). When high-risk patients were excluded, in-hospital mortality rates were very low at 0.4%. Figure 28 demonstrates the trend in in-hospital mortality rates over the previous 4 years. Low-risk PCI has consistently had very low mortality rates, while in-hospital mortality for patients with shock or out-of-hospital cardiac arrest have been trending down over the last 3 years.

Table 13: Unadjusted in-hospital mortality rates for selected patient groups

In-hospital Patient category Total mortality rate N N (%) All PCI patients* 9977 183 (1.8) All patients without STEMI or shock/intubated OHCA 7762 34 (0.4) STEMI patients 2166 132 (6.1) Shock/intubated OHCA patients 303 113 (37.3)

*Cases with multiple procedures were excluded to avoid mortality being counted more than once (n=15)

32 VCOR | 2016 Annual Report Figure 28: Comparative in-hospital mortality trends for different clinical presentation groups (2013-2016)

50 45 40 35 30 ) %

( 25

s e

s 20 a C 15 10 5 0 2013 2014 2015 2016 Ye ar

All PCI pae nts Al l (exc l. ST EMI & shock/OHCA ) STEMI pae nts Sh oc k/O HCA on ly

STEMI includes any STEMI patient with ACS onset within the last 7 days

In-hospital Bleeding

As an outcome measure, In-hospital major bleeding predicts adverse short and long-term outcomes, including prolonged hospital stay, increased risk of ischaemic events and higher mortality[10]. Utilising the international Bleeding Academic Research Consortium (BARC) standardised bleeding definitions for cardiovascular clinical trials[18], VCOR reports on major bleeding events (identified by BARC categories 3 and 5) to include bleeding that requires blood transfusion, cardiac tamponade, intracranial haemorrhage and/or any fatal bleeding.

The overall in-hospital major bleeding rate in 2016 was 1.0% (Table 14). High-risk groups for major bleeding included patients treated with ECMO (bleeding in 6 out of 11 ECMO patients, 55%), STEMI patients and females. Femoral arterial access was associated with a threefold increase in major bleeding compared with radial access (1.5% vs 0.5%). A comparison of in-hospital major bleeding rates among participating hospitals is shown in Figure 29.

Table 14: In-hospital major bleeding rates by sub-group

Sub-group N Major Bleeding Rate

Clinical Presentation N (%) STEMI 2160 46 (2.1) NSTE-ACS 3138 30 (1.0) Non-ACS 4668 19 (0.4) Gender N (%) Male 7633 56 (0.7) Female 2333 69 (1.7) Arterial Access Route N (%) Radial access 5508 30 (0.5) Femoral access 4444 65 (1.5) Brachial access 14 0 (0.0) Total 9966 95 (1.0)

Excludes ECMO cases (n=11) and cases with multiple procedures to avoid mortality being counted more than once (n=15)

VCOR | 2016 Annual Report 33 Figure 29: Rates of in-hospital major bleeding

8

Ho spitals 2 s ta ndard deviati ons 3 s ta ndard deviati ons 6 ) % (

g n i d e e

l 4 b

r o j a M

2

1. 0%

0

0 20 0 40 0 60 0 80 0 10 00 Number o f cases

Excludes ECMO cases (n=11) Site Y excluded due to low numbers (n<15)

In-hospital unplanned revascularisation In-hospital unplanned revascularisation refers to any unexpected revascularisation procedure (either PCI or CABG surgery) following the index PCI, and within the same admission. The measure includes unplanned procedures on coronary vessels other than the one initially treated with PCI. For 2016, the overall rate of in-hospital unplanned revascularisation was 0.7%. Figure 30 demonstrates that all participating hospitals had rates of unplanned revascularisation within control limits.

Figure 30: Rates of in-hospital unplanned revascularisation

8 Ho spitals 2 s ta ndard deviati ons 3 s ta ndard deviati ons

) 6 % (

n o i t a s i r a l u c s

a 4 v e r

d e n n a l p n U 2

0. 7%

0

0 20 0 40 0 60 0 80 0 10 00 Number o f cases

Site Y excluded due to low numbers (n<15)

34 VCOR | 2016 Annual Report Length of Stay For most elective PCI cases, length of stay is expected to be one day, whereas patients undergoing PCI for an acute coronary syndrome typically have multi-day hospital stays, usually related to the overall management of the underlying condition. Figure 31 shows the median length of stay (in days) was greatest for patients presenting with STEMI, decreasing as the acuity of the presentation decreased. A shorter median length of stay was observed among patients treated in private hospitals (3 days public vs 2 days private), likely reflecting the greater proportion of non-ACS cases treated in the private sector.

Figure 31: Length of stay by clinical presentation 0 1 8 ) s y a 6 d (

y a t s

f o

h t g 4 n e L 2 0

STEMI ( n=2166) NSTEMI ( n=2410) UAP ( n=728) Non-ACS (n=4673) Clinical p resenta tion

30-day risk-adjusted mortality

The 30-day risk adjusted mortality rate is one of the principal outcome measures used to assess the quality and safety of hospital care and is a crucial component for valid and meaningful comparisons of hospital performance. Despite well-recognised limits to the precision of risk-adjustment models, the technique of risk-adjustment helps level the playing field for inter-institutional comparisons. The metric essentially reflects how the actual (observed) outcomes for a hospital compare with outcomes that are expected for that hospital.

While the overall risk-adjusted 30-day mortality for the cohort was 2.8%, the real strength of the measure is in the comparison of the hospitals’ outcomes. Figure 32 demonstrates that all but one hospital had adjusted 30-day mortality rates within control limits. One hospital demonstrated special-cause variation as an outlier, with observed performance significantly different to its expected performance. In response VCOR followed its policy for management of outliers with appropriate notification to the hospital along with the offer of guidance and assistance with the hospital’s subsequent quality assurance activities.

VCOR | 2016 Annual Report 35 Figure 32: Risk-adjusted 30-day mortality

15 Ho spitals 2 s ta ndard deviati ons 3 s ta ndard deviati ons ) % (

y t i l

a 10 t r o m

y a d - 0 3

d e t s u j d a

- 5 k s i R 2. 8%

0

0 20 0 40 0 60 0 80 0 10 00 Number o f cases

Site Y excluded due to low numbers (n<15)

The current risk-adjustment model utilised by VCOR incorporates cardiogenic shock as one of the included clinical characteristics of the model. Yet, cardiogenic shock is generally associated with mortality rates much higher than other clinical presentations, and arguably is a special case that should be analysed separately to the rest of the cohort. Figure 33 demonstrates the 30-day mortality rate for patients with cardiogenic shock or out-of-hospital cardiac arrest. As expected, all hospitals demonstrated much higher mortality rates in this patient subgroup compared with an overall unadjusted 30-day mortality rate of 1.1% for patients without shock and out-of-hospital arrest. Some hospitals had very low numbers of cases and therefore comparisons of mortality rates across hospitals should be interpreted with caution.

Figure 33: 30-day mortality rates for cardiogenic shock and intubated OHCA patients by site

10 0 90 80 70 ) %

( 60

y t i l

a 50 t r o 40 M 30 20 10 0% 0 AC DEFGHI KLMOPRSTVX All Ho spital

Hospital A C D E F G H I K L M O P R S T V X All N 3 16 20 31 8 16 7 30 79 19 3 28 12 2 17 1 4 7 303

Hospitals B, J, N, Q, U, W, Y had no shock or intubated OHCA cases

36 VCOR | 2016 Annual Report 30-day Major Cardiac and Cerebrovascular Events (MACCE)

The composite endpoint of major adverse cardiac and/or cerebrovascular events (MACCE) includes all cases of death, new or recurrent myocardial infarction or stent thrombosis, target vessel revascularisation or stroke. The overall MACCE rate was 4.4% (Table 15). When the high acuity cases with cardiogenic shock or out-of-hospital cardiac arrest were excluded (n=303 cases), the 30-day MACCE rate dropped to 3.1%.

Table 15: 30-day MACCE rates

MACCE Component In-hospital events 30-day events * In-hospital NM (%AC) NC E(% C) omponent 30-day events* events N (%) N (%) Total mortality 183 (1.8) 218 (2.2) Myocardial infarction 63 (0.6) 101 (1.0) Stroke 34 (0.3) 50 (0.5) Definite stent thrombosis 16 (0.2) 39 (0.4) Probable stent thrombosis 9 (0.1) 15 (0.2) Target vessel revacularisation (TVR) ϯ 67 (0.7) 126 (1.3) Categories are not mutually exclusive *30-day events reported include in-hospital events ϯTVR refers to any ‘unplanned’ PCI or CABG revascularisation of the target vessel

A comparison of hospitals’ performance in relation to MACCE is shown in Figure 34 and demonstrates only common-cause variation among all the hospitals and no performance outliers.

Figure 34: 30-day MACCE

15

Ho spitals 2 s ta ndard d eviati ons 3 s ta ndard deviati ons

10 ) % (

E C C A M

5

3. 1%

0

0 20 0 40 0 60 0 80 0 10 00 Number o f cases

Excludes all shock/OHCA cases (n=303) Site Y excluded due to low numbers (n<15)

VCOR | 2016 Annual Report 37 30-day stent thrombosis Stent thrombosis was reported as a “definite” event (symptoms suggestive of an acute coronary syndrome and angiographic or pathologic confirmation of stent thrombosis) or a “probable” event (unexplained death within 30 days or target vessel myocardial infarction without angiographic confirmation of stent thrombosis). In 2016, the definite 30-day stent thrombosis rate was 0.4% (Table 15). There were no major differences in stent thrombosis rates among hospitals overall or when the public private sectors were compared.

30-day rehospitalisation At 30 days following PCI the overall rate of rehospitalisation was 13.7%, similar to the previous year. Readmission for cardiac causes occurred in 70% of cases, but only approximately one-third were unplanned (3.7% for the entire cohort, Table 16). The rate of unplanned readmission is being increasingly used as a metric of the performance or quality of hospital care or treatment and a comparison of this rate among participating hospitals is shown in Figure 35. Rates of unplanned cardiac readmission were similar among public and private hospitals.

Table 16: Rehospitalisation rates

All Patients Public Patients Private Patients Rehospitalisation Type (N=8308) (n=5205) (n=3103) N (%) N (%) N (%) Total readmissions 1141 (13.7) 632 (12.1) 509 (16.4) Non-cardiac readmissions 338 (4.0) 231 (4.4) 107 (3.4) Cardiac readmissions 803 (9.7) 401 (7.7) 402 (13.0) Unplanned cardiac readmissions 305 (3.7) 203 (3.9) 102 (3.3) Planned cardiac readmissions 498 (6.0) 198 (3.8) 300 (9.7)

Figure 35: 30-day unplanned cardiac rehospitalisation by hospital

20 18 16 14

) 12 % (

s

e 10 s a C 8

6 4

2 0% 0

Ho spital

38 VCOR | 2016 Annual Report Quality of Life Metrics Patients were requested to rate their perceived quality of life at 30-day follow-up in a series of questions based on a standardised measure, the EQ-5D[19]. Patients reported on their mobility, ability to perform usual domestic and personal care tasks, level of pain or discomfort and whether they experienced any anxiety or depression. The results are shown in Table 17. Patients were also asked to rate their own health status on a scale from 1-100, with 100 being the best a patient could remember ever feeling. Of the 44% of patients who responded, the median rating score for their own health status was 80 out of 100 (IQR 70, 90).

Table 17: Quality of life metrics for 2016 PCI cohort

PCI Cohort No problem Some problems Not asked (N=9992) N (%) N (%) N (%) Mobility 5971 (62.2) 753 (7.8) 2881 (30.0) Personal care 6397 (66.6) 331 (3.4) 2877 (30.0) Usual activities 5671 (59.0) 1053 (11.0) 2881 (30.0) Pain/discomfort 5886 (61.3) 819 (8.5) 2900 (30.2) Anxiety/depression 5021 (52.3) 896 (9.3) 3688 (38.4)

Patterns of responses to the EQ-5D were quite similar when patients were divided into age groups of <50 years, 50-75 years and >75 years. Interestingly, there was a higher rating of “moderate or extreme anxiety/depression” among the youngest patients (12.7%) compared with 50-75 year (9.6%) and >75 year groups (7.1%) (Figure 36).

Figure 36: EQ-5D responses to anxiety/depression by age group

100 90

80 42 38 39 70

) 60 97 % (

s 50 11 e s a

C 40 30 53 54 20 46 10 0 <5 0yrs 51-75yrs >7 5yrs Age group No anxiety/ de pr ession Mo derat e anxi ety/ depression Ex treme a nx iety/depressi on Not a sk ed

The quality of life responses were also analysed according to clinical presentation as it would be expected that the patients who presented with the highest acuity disease may have different responses to those patients who had stable disease. Overall, more complete response rates were noted for patients with non-ACS presentations (82% for non-ACS vs 61% for ACS). However, there were no major differences in each of the 5 quality of life categories assessed according to clinical presentation.

VCOR | 2016 Annual Report 39 Management of Acute ST-Elevation Myocardial Infarction (STEMI) in Regional Victoria (Early STEMI Management)

Background

A major challenge in Australian health care is ensuring patients in non-metropolitan regions with time critical illnesses have rapid availability of emergency specialist care. It is well recognized that patients who present with acute coronary syndrome in rural and regional areas face additional challenges related to timely treatment and definitive cardiology care[20]. The VCOR Early STEMI Management module was designed to monitor and report on the treatment of STEMI patients in non- metropolitan hospital settings, including aspects of pre-hospital care, initial emergency department management and timeliness to revascularisation and coronary angiography. VCOR developed this module because of the strong evidence base for effective treatment, well-developed standards of care and measurable process and outcome performance indicators.

Registry Module Activity

The Early STEMI Management module enrolls all patients with suspected STEMI who either present to the Emergency Department (ED) or are in-patients at the index rural or regional hospital, irrespective of whether they are deemed suitable for thrombolysis. The registry was modelled on the VCOR PCI module with a standard set of essential and epidemiologically sound variables collected. Data elements included details on reperfusion therapy, in-hospital clinical events, complications and clinical outcomes. Special focus was also given to data on inter-hospital transfers, as Victorian rural and regional health services typically identify delays in timely inter-hospital transfer to metropolitan hospitals as the single biggest obstacle to efficient and cost-effective treatment of STEMI.

The registry commenced activities in 2013, with four sites in the Hume and Gippsland regions recruited to pilot the module. The number of hospitals actively contributing to the registry has now expanded to nine, with the continued rollout of the module to other major regional centres across Victoria in the last 18 months (Table 18). The registry promotes regular engagement with its regional stakeholders through published bulletins and biannual data summaries. VCOR has also been visiting its regional stakeholders for direct engagement with local teams and incorporating the VCOR data into practice improvement for STEMI patients.

Data Linkage

For the first time in 2016, key data linkages and collaborations were established to provide a more comprehensive picture of the patient journey and outcomes in STEMI. The Early STEMI Management module was linked with the VCOR PCI module, tracking outcomes of non- metropolitan patients after transfer from their index hospital. Further data linkage was performed with the Ambulance Victoria (AV) pre-hospital database, providing important pre-hospital details and interventions.

40 VCOR | 2016 Annual Report Table 18: Participation of regional Victorian hospitals in Early STEMI Management module

Victorian Regional hospital Hospital 2013 2014 2015 2016 type

Albury Wodonga Health (Albury Campus) Public • Albury Wodonga Health (Wodonga Campus) Public • Bairnsdale Regional Health Service Public • • Bendigo Health Public • Central Gippsland Health Service (Sale) Public • • • Goulburn Valley Health (Shepparton) Public • • • • Latrobe Regional Health (Traralgon) Public • • • • Mildura Base Hospital Public • Northeast Health (Wangaratta) Public • • ° West Gippsland Healthcare group (Warragul) Public • • • • Wimmera Base Hospital (Horsham) Public •

Table Legend: • = contributing data; ° = engaged but not contributing

Patient Characteristics

In 2016, a total of 148 patients presented with suspected STEMI at participating hospitals. Data were collected from nine regional hospitals. Of the 148 patients, 41 (28%) were ineligible for thrombolysis. These included late presentation (n=15), significant comorbidities (n=4), uncertain diagnosis (n=5) or contra-indication to thrombolysis (n=17). Four patients were transferred for emergency primary PCI. The remaining 103 patients, all received appropriate thrombolytic therapy, either in the hospital ED (n=94) or through the pre-hospital thrombolysis scheme administered by Ambulance Victoria (n=9). This program has been rolling out in a sequential fashion across regional Victoria since 2014.

The average age for the 2016 regional STEMI cohort was 66 years, ranging from 31 to 95 years. Table 19 shows selected characteristics of the patients.

Table 19: Regional STEMI patient characteristics (2013-2016)

2014 2015 2016 Patient characteristic (N=64) (N=138) (N=148) Age – years (mean ±SD) 61 ±14 66 ±12 66 ±13.7 Presenting heart rate - BPM (mean ±SD) 79.4 ±21.3 79.1 ±24.8 81.5 ±22.4 % % % Gender – female 31.3 34.1 33.1 Pre-hospital thrombolysis (ambulance presentation only) 9.1 10.1 10.7 Site of infarction – anterior 34.4 40.6 33.8 Site of infarction – inferior 48.4 51.4 48.6 Site of infarction -posterior 6.3 2.2 6.8 Site of infarction –lateral - - 6.8 Site of infarction –intermediate - - 0.7 t characteristic 201Si4te(N o=f 6in4fa) rction – o2t0he1r5 ( N = 1 3 8 ) 2 0 1 6 ( N = 1 4 8 ) 10.9 5.1 3.4

Of the 148 patients, 84 (57%) were brought in by ambulance, and 54 (36%) of patients self-presented to ED. Four percent of STEMIs were current in-patients at regional centres, and approximately 3% of patients arrived from another treating hospital (inter-hospital transfer). While most patients (90%) were located within 50km of the treating hospital, almost 6% of patients travelled more than 75kms for treatment.

VCOR | 2016 Annual Report 41 Time Delays to Transfer

For all patients except those with in-hospital STEMI, the median time from pain onset to first medical contact (patient delay) was 120 minutes (IQR: 37, 268). The median time from first medical contact to hospital arrival (pre-hospital delay) was 66 minutes (IQR: 37, 85). For patients who travelled less than 25kms to hospital, the median pre-hospital delay was shorter at 53 minutes (IQR: 33, 75), for those travelling between 26 and 50kms, the median time was 90 minutes (IQR: 79, 103). Patients travelling more than 50kms, the median time was 106 minutes (IQR: 73, 196). Time intervals related to ambulance calls are noted in Figure 37.

Figure 37: Ambulance time intervals for regional STEMI patients

Symptom onset to Ambulance call Ambulance arrival receive to to hospital arrival ambulance call Ambulance arrival Median = 52 mins Median = 90mins Median = 12mins (IQR: 36, 71) (IQR 21, 263) (IQR 10, 24)

In-hospital Process Times (Arrival to ECG time and Door-to-Needle Time) The efficient and timely delivery of reperfusion therapy is a key performance outcome and can be evaluated with a number of specific process measures. These include the time to first ECG and the time taken from hospital presentation to administration of thrombolytic drug, known as the door-to-needle time. Across all sites in the 2016 cohort, the median time to first ECG was at the recommended benchmark of 10 minutes (Figures 38).

Figure 38: Time from arrival to first ECG time for regional STEMI patients by hospital

Target ti me t o f irst E CG ( 10 mins) 0 8 ) s 0 n i 6 m (

G C E

t s r i f

o 0 t

4 e m i t

n a i d e M 0 2 0

S1 S2 S5 S6 S7 S8 Ho spital

Hospitals S3, S4 and S9 not included due to low case numbers (n<5)

42 VCOR | 2016 Annual Report The overall median door-to-needle time for VCOR regional STEMI sites was 39 minutes (IQR: 27, 69). The ideal target door-to-needle time (DTN), recommended by Australian National Heart Foundation/Cardiac Society of Australia and New Zealand Guidelines is less than 30 minutes[13]. This ideal median target of 30 minutes was not achieved by any of the sites in 2016 (Figure 39).

Figure 39: Door-to-needle times for regional STEMI patients by hospital

0 Target D TN ti me ( 30 m ins) 0 2 ) 0 s 5 n 1 i m (

e m i t

e l d e 0 e 0 n - 1 o t - r o o d

n a i d e 0 M 5 0

S1 S2 S3 S5 S6 S7 S8 Ho spital

Hospitals S4 and S9 not included due to low case numbers (n<5)

When hospital performance was analysed by compliance rates for door-to-needle times ≤30 minutes, 32% cases were managed within this benchmark rate. The proportion of compliant cases rose to 68% when a door-to-needle time delay of ≤60 minutes was applied (Figure 40).

Figure 40: Proportion achieving door-to-needle times within 30 and 60 minutes (regional STEMI patients)

100 90 80 70 60

) 50 45 % (

s

e 40 s a

C 30 20 10 0% 0 S1 S2 S3 S4 S5 S6 S7 S8 S9 All Hos pital Wi thin 30 mins Wi thin 60 mins

Hospitals S4 and S9 had low case numbers (n<5)

VCOR | 2016 Annual Report 43 As with door-to balloon times in patients undergoing PCI, door-to-needle times (DTN) with thrombolysis also improved when pre- hospital notification of the arriving STEMI patient was received from ambulance services. The median door-to-needle time with pre-hospital notification was 24 minutes (IQR: 16, 56) among the 18 patients who were triaged in the field. 10 of the 18 cases with pre-hospital notification achieved a median door-to-needle time within 30 minutes. Table 20 shows DTN for all cases thrombolysed at the index hospital, including self-presenters and ambulance arrivals with and without pre-hospital notification.

Table 20: Door-to-needle times for pre-hospital notification (PHN) and no PHN (regional STEMI patients)

door-to-needle Median door-to-needle Median door-to-needle Hospital Total cases Median time (all cases) time (with PHN) time (no PHN) N Mins (IQR) Mins (IQR) Mins (IQR) S1 10 43 (29, 308) - 43 (31, 754) S2 15 34 (28, 53) - 34 (31, 61) S3 5 32 (15, 73) - 54 (32, 92) S4 2 - - - S5 14 47 (37, 109) 39 (19, 59) 61 (38, 148) S6 9 52 (30, 104) - 52 (30, 104) S7 18 46 (23, 90) - 48 (35, 90) S8 20 33 (23, 43) 28 (12, 70) 35 (23, 43) S9 1 - - - All sites 94 39 (27, 69) 24 (16, 56) 41 (30, 70)

)Hospitals S1, S2, S3, S4, S6, S7, S9 had low numbers (n<3)

A comparison of the system delay (comprising pre-hospital delay plus door-to-needle time) for the participating hospitals is shown in Figure 41. The median system delay for the entire cohort was 64 minutes, substantially longer than the median door-to-needle time of 39 minutes. This measure is arguably a better performance metric than the door-to-needle time, as it emphasises the urgency of commencing treatment the moment the patient comes into contact with the medical system. International guidelines are now recommending ideal time delays from first medical contact to thrombolysis of <30 minutes[21], underscoring the importance of continued development of pre-hospital thrombolysis programs to shorten first medical contact to reperfusion.

Figure 41: Overall system delay times (regional STEMI patients)

S1 38 10 30

S2 43 8 27

S3 41 5 25 l

a S5 82 10 43 t i p s

o S6 5 10 47 H

S7 34 10 39

S8 70 5 26

02 0406080100120140 Ti me (mins) Me dian Pre-hospital delay Me dian Tim e to first ECG Me dian Tim e EC G to Lysis

Hospitals S4 and S9 excluded due to low case numbers (n<5)

44 VCOR | 2016 Annual Report Adjunctive Therapies

Australian guidelines provide strong recommendations for antithrombotic therapy (unfractionated heparin or low molecular weight heparin), and anti-platelet therapy with clopidogrel following thrombolysis[13]. Figure 42 shows very high rates of both antithrombotic therapy (96%) and clopidogrel use (78%) among patients.

Figure 42: Treatment and outcomes: adjunctive therapies (regional STEMI patients)

100 90 80 70 60 ) % (

50 s e s

a 40 C 30 20 10 0 S1 S2 S3 S4 S5 S6 S7 S8 S9 All Hos pital Cl opidogrel Fu ll d ose L MHW/H epar in

Hospitals S4 and S9 had low case numbers (n<5)

In-hospital Outcomes and Transfer Rates

The mean unadjusted in-hospital mortality rate for the nine participating hospitals was 6.1%, comparable to other international registries of STEMI [22, 23] (Table 21). Cardiogenic shock (CS) occurred in 8.8% of patients, and was associated with an in-hospital mortality rate of 54% (7 deaths in 13 patients with CS). Major bleeding occurred in three cases (2%) but there were no cases of stroke among treated patients during their stay at the index hospital.

Table 21: In-hospital outcomes for regional STEMI patients

Hospital Total cases Mortality Cardiogenicshock M y o Ccaarrddioiagleren-icin f a r c ti o n MyocardMiaal j o r b l eeding Stroke Hospital Total cases Mortality Major bleeding Stroke Shock re-infarction N % % % % % S1 19 5.3 10.5 0.0 0.0 0.0 S2 18 0.0 5.6 0.0 0.0 0.0 S3 6 0.0 0.0 0.0 0.0 0.0 S4 4 0.0 0.0 0.0 0.0 0.0 S5 38 15.8 18.4 2.6 2.6 0.0 S6 13 0.0 7.7 7.7 0.0 0.0 S7 21 0.0 4.8 0.0 4.8 0.0 S8 20 5.0 5.0 5.0 0.0 0.0 All sites 1S498 6 . 1 8 . 8 9 2 . 7 112..00 0 0.0 11.0 11.0 0.0 All sites 148 6.1 8.8 2.7 2.0 0.0

VCOR | 2016 Annual Report 45 Importantly, most patients (95%) were transferred to a PCI capable hospital within 24 hours of thrombolytic therapy. The median time from referral request to the actual transfer from VCOR regional STEMI sites to metropolitan PCI hospitals was 2.3 hours (IQR: 1.4, 4.2) (Figure 43).

Figure 43: Patient transfer times to metro VCOR hospital (regional STEMI patients)

24hrs 13-24hrs 5% 6%

7-12hrs 9%

6 hrs 80%

Revascularisation Rates

For the first time in 2016, VCOR was able to link patient data across its modules. Therefore providing crucial follow-up and clinical outcome data to regional STEMI patients who underwent subsequent PCI. Probabilistic matching indicated that of the 148 regional patients treated for STEMI, 126 patients were transferred to a PCI centre (85%). Of these patients, 68% had PCI and 9% underwent CABG (Figure 44).

Figure 44: Regional STEMI module inter-hospital transfer and revascularisation rates

Ɵ Presen ng regionally with suspected STEMI N=148

Ɵ Transferred to ter ary referral centre N=126 (85%) •VCOR (PCI) Underwent PCI n=86 (68%) Underwent CABG n=11 (9%)

46 VCOR | 2016 Annual Report Further analysis of the 86 patients who underwent subsequent PCI included the 4 patients who were transferred urgently for primary PCI. These 4 patients had no in-hospital complications, with three patients discharged to home and one to inpatient rehabilitation. Twenty patients were transferred for a rescue PCI, in which thrombolysis had not achieved reperfusion of the occluded artery. For these patients, urgent transfer and rescue PCI are indicated. The median time from thrombolysis to PCI was 5 hours (IQR: 4, 6) for this group. The in-patient mortality rate for the rescue PCI cohort was 10%.

The remaining patients were transferred in a stable state and underwent pharmaco-invasive PCI (PCI within 24 hours of successful thrombolysis). The median time from thrombolysis to PCI was 22 hours (IQR: 9, 34). There were no deaths in this group (Table 22).

Table 22: Outcomes for regional STEMI patients by reperfusion strategy

Primary Pharmaco-invasive Rescue O u t c o m e s PCI Cohort PCI Cohort PCI Cohort (N=4) (N=62) (N=20) Median time from thrombolysis to PCI – hrs (IQR) n/a 22 hrs (9, 34) 5 hrs (4, 6) (%) (%) (%) Lesion successfully treated 100 98.8 95.5 In-hospital mortality 0.0 0.0 10.0 In-hospital unplanned revascularisation 0.0 0.0 0.0 In-hospital major bleeding 0.0 4.8 0.0 30-day mortality 0.0 1.6 10.0 30-day unplanned revascularisation 0.0 1.6 0.0 30-day MACCE 0.0 4.8 10.0

VCOR | 2016 Annual Report 47 Heart Failure (HF) Snapshot

Background

Heart failure (HF) represents a major public health problem associated with high mortality, frequent hospitalisation and major utilisation of health care costs [24]. Thus, optimal utilisation of resources to impact the public burden of heart failure is a significant health system priority. This requires a clear understanding of the epidemiology of the condition, the scope of currently available management strategies and demographic characteristics that predict rehospitalisation. Although considerable data exist from international sources, there are only limited data on the epidemiology and public burden of heart failure in Victoria.

In 2014, VCOR received funding to commence a pilot project related to heart failure patients in Victoria. The overall aim was to improve the safety and quality of care provided to these patients. The design involved the periodic collection and analysis of key clinical information from individual healthcare encounters to facilitate benchmarking of performance and improvement in the quality of health care services. The pilot project, entitled HF-Snapshot, enrolled consecutive patients at participating health services for a limited period of time (one month), in order to obtain a “snapshot” of heart failure-related treatment and outcomes. The pilot phase was successfully completed in 2014 and resulted in the development of a functional minimum dataset for a heart failure clinical quality registry.

Registry Module Activity

In 2016, work proceeded on all aspects of the project, including the updating of all data elements. The number of participating sites increased to 16 hospitals. Data collection activities for 2016 were completed by year’s end. Overall, there was strong engagement and interest shown by hospitals in the HF-Snapshot project.

All patients admitted to hospital with acute decompensated heart failure over a one month period during June-July 2016 were enrolled into the 2016 snapshot (Table 23) and followed-up at 30 days post-discharge.

Table 23: Participation of hospitals in HF-Snapshot (2014-2016)

Hospital Hospital type 2014 2015 2016 Alfred Hospital, The Public • • • Austin Hospital Public • • • Bairnsdale Regional Health Service Public • • Bendigo Hospital Public • • • Box Hill Hospital Public • • Central Gippsland Health Service (Sale) Public • • Dandenong (Monash Health) Public • Epworth Hospital Richmond Private • Footscray Hospital Public • • • Frankston Hospital Public • • Latrobe Regional Hospital (Traralgon) Public • Northern Hospital, The Public • • MonashHeart (Monash Medical Centre Clayton) Public • • • Royal Melbourne Hospital, The Public • • Public • St Vincent’s Hospital Melbourne Public • • • University Hospital, Geelong Public • • Table Legend: • = contributing data

48 VCOR | 2016 Annual Report Patient Characteristics

The 16 participating hospitals enrolled 456 patients admitted to hospital with acute decompensated heart failure. The majority of the patients were male (58%) with a median age of 76 years. The most common co-morbidities were atrial fibrillation (54%), diabetes (47%), anaemia (36%), history of angina (34%), chronic obstructive pulmonary disease (COPD) /asthma (31%), and moderate chronic kidney disease (31%) (Table 24).

Table 24: HF-Snapshot patient characteristics (2015-2016)

2015 cohort 2016 cohort Patient characteristics (N=289) (N=456) Age – years (Mean ±SD) 77 ±13 76 ± 14 % % Gender – female 42.9 41.9 Cardiovascular Medical History % % Treated diabetes 39.1 46.5 Cerebrovascular disease 15.6 19.7 History of angina 41.2 33.8 History of MI 28.7 30.0 History of heart failure 65.7 78.9 Arrhythmia 54.3 54.3 History of PCI or CABG 30.1 31.1 CIED therapy 22.1 23.0 Non-Cardiovascular Medical History % % COPD/Asthma 34.9 30.7 Obstructive sleep apnoea 15.3 11.8 Anaemia 24.2 35.7 Iron deficiency 12.5 25.9 Depression 16.3 18.9 Liver disease -Mild 4.9 2.6 Liver disease -Moderate/severe 3.1 2.2 Chronic Kidney disease -Mild 13.8 20.0 Chronic Kidney disease -Moderate 31.5 31.4 Chronic Kidney disease -Severe/dialysis 13.5 14.3 Dementia 8.7 6.8

For patients admitted to hospital with acute decompensated heart failure, 61% had experienced a previous hospitalisation for heart failure. Ischaemic cardiomyopathy was present in 37% of these patients and of these, 31% had a history of prior CABG and/or PCI. Hypertensive heart failure occurred in 22% of patients. Clinical Presentation

Nearly all the patients (85%) were admitted to hospital through the Emergency Department. Overall, 49% were admitted under General Medicine and 38% under Cardiology (Table 25). Three hospitals had a dedicated heart failure unit. Yet even in those hospitals, the majority of heart failure patients were admitted under General Medicine. Medication non-adherence was considered the precipitant for admission in 8% of patients, similar to 2015 (7%). In contrast, fewer patients were admitted with fluid overload as a precipitant in 2016 (33% in 2016 vs 48% in 2015). Other precipitants for admission are shown in Table 25.

VCOR | 2016 Annual Report 49 Table 25: HF-Snapshot clinical presentation (2015-2016)

2015 cohort 2016 cohort Clinical presentation (N=289) (N=456) % % Admission location – Emergency Department 90.3 84.6 Admitting speciality % % Heart failure unit 8.0 6.4 Cardiology 29.8 37.5 General medicine 52.2 49.3 Gerontology 3.1 2.4 Precipitant for admission % % Ischemia 10.0 9.9 Medication-non-compliance 7.3 8.3 Medication-precipitating drugs 1.4 1.8 Shortness of breath 57.8 88.2 Rhythm disturbance 14.9 9.9 Infection 20.8 18.2 Fluid overload 48.4 32.7

On admission, 41% of patients were in sinus rhythm and 42% were in atrial fibrillation. At discharge, the proportion of patients in sinus rhythm had increased to 48%. There was also an improvement in patients’ functional status (as assessed by the New York Heart Association (NYHA) class) by the time of discharge (Figure 45). The greatest improvement in functional class was in the sickest patient group on presentation (NYHA Class IV). A reasonable proportion of patients did not have their functional class adequately documented at admission and or discharge.

Figure 45: New York Heart Association (NYHA) class rates on admission and discharge during HF-Snapshot

90

80

70

60

) 50 % (

s e

s 40 a C 30

20

10

0 Cl as s I Cl as s I I Cl as s II I Cl as s I VNot Do cumented NY HA class Ad mission Dis charg e

Overall, 63% of patients had left ventricular dysfunction. Of these, 46% of patients were diagnosed with predominantly systolic dysfunction (heart failure with reduced ejection fraction, HFrEF) and 14% had predominantly diastolic dysfunction (heart failure with preserved ejection fraction, HFpEF) on echocardiography.

50 VCOR | 2016 Annual Report During their hospital admission, 84% of patients were prescribed intravenous frusemide and 87% were prescribed oral diuretics. Medications prescribed on admission and discharge for the entire heart failure cohort are listed in Table 26. These rates do not take into account the type of heart failure the patient had (HFrEF vs HFpEF) nor the contraindications associated with prescribing these medications such as ACE inhibitors (ACEIs), beta-adrenergic blockers, angiotensin receptor blockers (ARBs), ivabradine and/or aldosterone antagonists in heart failure with reduced ejection fraction (HFrEF).

Table 26: Medications prescribed at admission and discharge for all patients during HF-Snapshot

At admission At discharge Medications (N=456) (N=429)* % % AACE Inhibitor 35.1 42 Beta Blocker 60.7 71.8 ARB 19.1 20.5 Ivabradine 2.9 3.3 Calcium channel antagonist 19.3 16.6 Digitalis 15.4 16.6 Nitrate 14.7 16.1 Other vasodilator 5.5 5.8 Antiarrhythmic 8.8 7.9 Lipid lowering agent 53.9 58.5 Anticlotting agents % % Antiplatelet 45 50.8 Anticoagulant 38.4 43.1 Diuretic Agents % % Aldosterone antagonist 21.1 34.3 Loop diuretic 70.6 93.7 Thiazide diuretic 6.4 9.1

*In-hospital mortality cases not included (n=27)

When adjusting for contraindications and only including patients diagnosed with HFrEF, the prescribing of beta-adrenergic blockers, ACEI/ARBs and aldosterone antagonists all increased from time of admission to time of discharge (Figures 46 and 47).

Figure 46: Prescribing of beta-adrenergic blockers in HFrEF cohort with heart rate >60 BPM on admission and discharge during HF-Snapshot 80

70

60

50 ) % (

s

e 40 s a C 30

20

10

0 On admis sion On disch arge

VCOR | 2016 Annual Report 51 Figure 47: Prescribing of ACEI/ARB in HFrEF with eGFR >60 on admission and discharge during HF-Snapshot

10 0 90 80 70

) 60 % (

s

e 50 s a C 40 30 20 10 0 On admission On di sch arge AC E/AR B Al dos ter one ant agoni st

Median length of hospital stay was 6 days (IQR: 3, 9). Overall, 72% were discharged to home and a further 7.5% transferred to rehabilitation unit or hospital (Table 27).

Table 27: Discharge destination during HF-Snapshot

2015 cohort 2016 cohort Discharge Destination N=289 N=456 % % Home 73.4 72.1 Hospital In The Home 2.1 2.2 Rehabilitation Unit / Hospital 8.7 7.5 Nursing Home / Aged Care Facility 6.2 6.8 Local or Referring Hospital 2.8 2.9 Palliative Care 2.1 0.4 . Tertiary Referral Centre 0.0 2.2 In-hospital Mortality 4.8 5.9

The overall unadjusted in-hospital mortality for the 2016 cohort was 6%. Figure 48 demonstrates comparative rates of in- hospital mortality among hospitals. There were no performance outliers for in-hospital mortality and no apparent relationship between mortality and hospital volume.

52 VCOR | 2016 Annual Report Figure 48: In-hospital all-cause mortality during HF-Snapshot

Ho spitals 80 2 s ta ndard deviati ons 3 s ta ndard deviati ons

) 60 % (

y t i l a t r o m

l a

t 40 i p s o h - n I

20

5. 9%

0

0 20 40 60 80 Number o f cases

Transitional Care after Discharge

For the entire cohort, 63% of patients had an outpatient appointment scheduled at discharge, and by 2 weeks, 27% had been seen. Referrals were either to a public sector outpatient clinic or to a private cardiologist/clinic. For public sector outpatients, 49% of patients were seen within 30 days of discharge. For patients followed up privately, 25% were reviewed within 14 days and 39% were seen within 30 days of discharge.

Referrals to a heart failure program were low at 34% (Figure 49). Only 6% of patients had a pharmacist medication review arranged post-discharge and 10% were referred to a heart failure exercise program (Figure 49). While these rates appear below ideal levels, there may be an under-estimation of “appropriate” referral rates. The HF-Snapshot was unable to identify those patients who were ineligible or unsuitable for referral to a heart failure program, pharmacist medication review or exercise program, as different sites had varying inclusion and exclusion criteria. These results were similar among hospitals and have been fed back to the participating sites. Deficiencies in transitional care arrangements post-discharge identified in this report highlight the need for improvements in processes relating to timely referrals and outpatient appointments.

Figure 49: Discharge planning during HF-Snapshot

40

35

30

25 ) % ( 20 s e s a

C 15

10

5

0 He art Tx / He art Failure He art Failure Car diac R ehab Phar mac ist LV AD Ex erc is e Pr ogram Medic ao n Pr ogram Re view

VCOR | 2016 Annual Report 53 Outcome measures

The unadjusted 30-day mortality rate was 9.9% for the overall cohort (Figure 50). This was similar to 2015 (9.3%), and comparable to international 30-day mortality rates[25]. The 30-day mortality rate was higher in patients diagnosed with HFrEF (8.9%) compared with HFpEF (4.8%). All hospitals were within control limits for 30-day mortality (Figure 51).

Figure 50: Overall outcome measures during HF-Snapshot 2016

40

35

30

) 25 % (

s

e 20 s a C 15

10

5

0 In Ho spital Mortali ty 30 Day Mor tal ity 30 Day Re ad mis sion

2015 2016

Figure 51: Unadjusted 30-day mortality during HF-Snapshot 2016

10 0 Ho spitals 2 s ta ndard deviati ons 3 s ta ndard deviati ons 80 ) % (

y 60 t i l a t r o m

y a d - 40 0 3

20

9. 9%

0

0 20 40 60 80 Number o f cases

At 30 days post-discharge, the all-cause readmission rate was 21% (Figure 52) - lower than in 2015 (26%) and comparative international rates of 25% [25] (Figure 50). One potential factor in this improved readmission rate was a separate program, undertaken during this time, aimed at targeting readmissions for heart failure. The Victorian Cardiac Clinical Network funded individual projects in 7 health services involved in HF-Snapshot to implement models of care aimed at reducing heart failure readmissions. All the projects demonstrated an improvement in short-term patient outcomes and this likely positively influenced 30-day readmission rates.

54 VCOR | 2016 Annual Report Figure 52: 30-day all-cause hospital readmission during HF-Snapshot 2016

10 0 Ho spitals 2 s ta ndard deviati ons 3 s ta ndard deviato ns 80 ) %

( 60

n o i s s i m d a e 40 R

21 .2% 20

0

0 20 40 60 80 Number o f cases

For those patients readmitted to hospital, the median time to readmission was 14 days (IQR: 7, 22) (Figure 53). Of the patients readmitted, 40% were readmitted within 10 days of discharge.

Figure 53: Time to readmission by hospital for HF-Snapshot 2016 0 3 ) s y a d ( 0

2 n o i s i m d a e r

o t

e m i t

n 0 a 1 i d e M 0

H1 H2 H3 H4 H5 H6 H7 H8 H9 H1 0 H1 2 H1 3 H1 4 H1 5 H1 6 H1 7 Ho spital

HF-Snapshot future work In 2017, another snapshot will be conducted from mid-May to June with planned 3-month follow-up across 16 hospitals. The follow-up period has been extended to provide a comprehensive perspective of inpatient and transitional care in a cohort of patients hospitalised with acute decompensated heart failure and known for its high burden of disease. It is hoped that there will be further improvements in readmissions, mortality and pharmacotherapy compared with 2016 results.

VCOR | 2016 Annual Report 55 Future Directions

In many respects, the future directions of VCOR will be strongly influenced by the culture of increased vigilance of hospital performance and the strengthening of safety and quality for patients that has emerged at a state level in Victoria in 2016. This is on a background of the recent report, Targeting Zero: Supporting the Victorian hospital system that proffered a comprehensive set of recommendations to ensure delivery of safe and continuously improving care by our hospital system. It complements the blueprint for the redesign, service and infrastructure for Victoria’s cardiac system, released by the Department of Health and Human Services that sets out the framework for these service improvements. Safer Care Victoria, together with the Cardiac Clinical Network, the Victorian Agency for Health Information and the Department of Health and Human Services will help implement these changes.

VCOR is looking forward to a busy and productive year ahead, as a principal tool for data collection, performance monitoring and benchmarking, and as an enabler of the identification of potential shortcomings in care.

From a workflow standpoint, VCOR has been focussing on a comprehensive revision of all the data elements throughout 2016, and version 2.0 of our PCI database will be released in 2017. Our robust auditing activities will continue to ensure the highest quality data are collected. VCOR will continue to expand its modules of early STEMI management and acute heart failure management with engagement of further sites and ongoing support for health services.

The 2017 year will also see an exciting addition to VCOR with the impending commencement of a fourth module to monitor the safety and quality of implantation of cardiac implantable electronic devices (CIED). In an initial pilot phase to commence in July 2017, five health services will participate in a registry module that both tracks contact details of the patient and device and clinical quality data to assess performance and outcomes of the patients and their devices as well as the health services that manage them.

VCOR will continue to develop linkages with large administrative datasets, organisations and other registries to enhance the quality and breadth of its data, while at the same time work collaboratively to the mutual benefit of all participants. The registry has also established links with researchers both within its network and with a number of external academic institutions in areas that include quality of life metrics, patient-related outcomes, epidemiological aspects of cardiac disease and health economics.

The activities of VCOR may be diverse, but its aim remains sharply focussed on ensuring that Victorian patients with cardiac conditions receive the highest quality care and the best outcomes possible irrespective of their location, insurance status or healthcare provider. Through the strong and ongoing support of Victoria’s cardiovascular clinicians in the medical, nursing and allied health spheres and committed involvement of other key stakeholders including funders, health providers, hospital executives and patient advocates, VCOR looks ahead to a productive future in the coming years.

56 VCOR | 2016 Annual Report Glossary

ACEI Angiotensin-Converting-Enzyme Inhibitors ACS Acute Coronary Syndrome ANZSCTS Australian and New Zealand Society of Cardiac and Thoracic Surgeons ARB Angiotensin Receptor Blockers ARC Academic Research Consortium BARC British Academic Research Consortium BB Beta-adrenergic Blockers BMS Bare Metal Stent BPM Beats Per Minute BVS Bio-resorbable Vascular Scaffold CABG Coronary Artery Bypass Graft CSANZ Cardiac Society of Australia and New Zealand CTO Chronic Total Occlusion DEPM Department of Epidemiology & Preventive Medicine DES Drug Eluting Stent DTB Door-to-balloon DTN Door-to-needle Time ECG Electrocardiograph ECMO Extracorporeal Membrane Oxygenation ED Emergency Department eGFR Estimated Glomerular Filtration Rate HF Heart Failure HFpEF Heart Failure with Preserved Ejection Fraction HFrEF Heart Failure with Reduced Ejection Fraction IABP Intra-aortic Balloon Pump IQR Inter Quartile Range KPI Key Performance Indicator LMHW Low Molecular Heparin Weight MACCE Major Adverse Cardiac & Cerebrovascular Event NHMRC National Health & Medical Research Council NSTE-ACS Non-ST Elevation Acute Coronary Syndrome NSTEMI Non-ST Elevation Myocardial Infarction NYHA New York Heart Association OHCA Out of Hospital Cardiac Arrest PCI Percutaneous Coronary Intervention POBA Plain Old Balloon Angioplasty RI Renal Impairment SD Standard Deviation STEMI ST-Elevation Myocardial Infarction TVR Target Vessel Revascularisation UAP Unstable Angina Pectoris VCOR Victorian Cardiac Outcomes Registry

VCOR | 2016 Annual Report 57 Publications and presentations in 2016

Asrar ul Haq, M, Tsay, IM, Dinh, DT, Brennan, A, Clark, D, Cox, N, Harper, R, Nadurata, V, Andrianopoulos, N, Reid, C, Duffy, SJ, Lefkovits, J & van Gaal, WJ. Prevalence and outcomes of trans-radial access for percutaneous coronary intervention in contemporary practise. International Journal of Cardiology. 2016; 221 , 264-268.

Brennan, A., Dinh, D., Lefkovits, J., Stub, D., Brien, R., Carruthers, H., Andrianopoulos, N., Duffy, S., Clark, D., Hiew, C., New, G., Warren, R., McNeil, J., Reid., C. Ensuring the Quality in Clinical Quality Registries -Design and Implementation of Rigorous Audit Activities in the Victorian Cardiac Outcomes Registry (VCOR). 64th Cardiac Society of Australia and New Zealand ASM, Adelaide. Heart, Lung & Circulation 2016; 25 (S2), S174.

Dinh, DT, Brennan, A, Stub, D, Reid, C & Lefkovits, J. (2016). Deep Dive into Treatment Delays for PCI for STEMI: The Victorian Cardiac Outcomes Registry (VCOR). Presented at: NHMRC Symposium on Research Translation; 23 November 2016. Melbourne, Vic, Australia.

Driscoll A, Brennan A, Dinh D, Lefkovits J, Hare DL, Carruthers H, Hopper I, Kaye D, Neil C, Prior D, Reid C. The Victorian Cardiac Outcomes Registry (VCOR) Heart Failure Snapshot. 64th Cardiac Society of Australia and New Zealand ASM, Adelaide. Heart, Lung & Circulation 2016; 25 (S2) , 123.

Lefkovits, J, Brennan, AL, Dinh, DT, Stub, D, Brien, R, vanGaal, W, Haikerwal, D, Cox, N, Duffy, SJ, Clark, D, Nadarajah, N, Harper, R, Thompson, E, Reid, C. Deep dive into hospital performance for PCI for acute STEMI: The Victorian Cardiac Outcomes Registry (VCOR). 64th Cardiac Society of Australia and New Zealand ASM, Adelaide. Heart, Lung & Circulation 2016; 25 (S2), S39.

Stub, D, Brennan, A, Dinh, D, Brien, R, Reid, C, Gutman, J, Toogood, G, Nadurata, V, Oqueli, E, Hengel, C, Sapontis, J., Harper, R., Lefkovits, J. The Victorian Cardiac Outcomes Registry (VCOR): Guiding high quality of PCI outcomes in Victoria. 64th Cardiac Society of Australia and New Zealand ASM, Adelaide. Heart, Lung & Circulation 2016; 25 (S2) , 203.

Stub, D, Brennan, A, Dinh, D, Brien, R, Nadurata, V, Amerena, J, Forge, B, Ibrahim, T, Krones, R, Connor, H, Wright, A, Young, T, Reid, C, Lefkovits, J. Measuring quality of early care and clinical outcomes for STEMI patients in Victoria regional hospitals utilising linkage across VCOR STEMI and PCI registries. 64th Cardiac Society of Australia and New Zealand ASM, Adelaide. Heart, Lung & Circulation 2016; 25 (S2), S51.

58 VCOR | 2016 Annual Report VCOR Personnel

VCOR Custodian Professor Christopher Reid

VCOR Project Co-Leads Professor Christopher Reid A/Prof Jeffrey Lefkovits Professor Richard Harper Ms Angela Brennan Professor John McNeil

VCOR Clinical Director A/Prof Jeffrey Lefkovits

Heart Failure Clinical Lead Prof Andrea Driscoll

Early STEMI Management Clinical Lead Dr Dion Stub

VCOR Site Leads (PCI Hospitals) Dr Stephen Duffy Alfred Hospital, The Dr David Clark Austin Hospital, The Prof Ernesto Oqueli Ballarat Base Hospital Dr Voltaire Nadurata Bendigo Hospital Dr Jennifer Cooke Box Hill Hospital A/Prof Jeffrey Lefkovits Cabrini Hospital Malvern A/Prof Ron Dick Epworth Hospital Richmond A/Prof Ron Dick Epworth Hospital Eastern A/Prof Ron Dick Epworth Eastern, Geelong Dr Geoff Toogood Frankston Hospital A/Prof Nicholas Cox Footscray Hospital A/Prof John Amerena Geelong Private Hospital Dr James Sapontis / Dr Robert Gooley Jessie McPherson Private Hospital Dr Michael Leung Linacre Private Hospital Dr Michael Rowe Knox Private Hospital A/Prof William van Gaal Northern Hospital, The Dr James Sapontis / Dr Robert Gooley MonashHeart (Monash Medical Centre Clayton) Dr Roderic Warren Melbourne Private Hospital Dr Olivier van den Brink Peninsula Private Hospital A/Prof Leeanne Grigg Royal Melbourne Hospital, The Dr Chris Hengel St John of God (Ballarat) Dr Nimalan Nadarajah St John of God (Bendigo) Dr Martin Sebastian St John of God (Geelong) A/Prof Jack Gutman St Vincent’s Private Hospital A/Prof Andrew MacIsaac St Vincent’s Hospital Melbourne A/Prof Nicholas Cox Sunshine Hospital Dr Chin Hiew University Hospital, Geelong Dr Jodie-Ann Senior Valley Private Hospital, The Dr Mark Horrigan Warringal Private Hospital Dr Deepak Haikerwal Western Private Hospital

VCOR | 2016 Annual Report 59 VCOR Site Leads (Early STEMI Management Hospitals) Dr Franz Eversheim Albury Wodonga Health (Albury Campus) Dr Franz Eversheim Albury Wodonga Health (Wodonga Campus) Dr Voltaire Nadurata Bendigo Hospital Dr Howard Conner Central Gippsland Health Service (Sale) Dr Tunde Ibrahim Goulburn Valley Health (Shepparton) Dr Alistair Wright & Dr Tony Chan Latrobe Regional Health (Traralgon) Dr Mark Wadsworth Mildura Base Hospital Dr Robert Krones Northeast Health (Wangaratta) Dr Brett Forge West Gippsland Healthcare group (Warragul) Dr Sanath Weerakkody Wimmera Base Hospital (Horsham)

VCOR Site Leads (HF-Snapshot Hospitals) Prof David Kaye Alfred Hospital, The Prof Andrea Driscoll & Prof David Hare Austin Hospital, The Dr Justin Mariani & Ms Debbie Gash Bairnsdale Regional Health Service Dr Voltaire Nadurata Bendigo Hospital Dr Jennifer Cooke Box Hill Hospital Dr Howard Conner Central Gippsland Health Service (Sale) Dr Siobhan Lockwood & Prof James Cameron Dandenong Hospital (Monash Health) Dr Hendrik Zimmet Epworth Hospital Richmond Dr Chris Neil Footscray Hospital Dr Geoff Toogood & Dr Phillip Carrillo Frankston Hospital Dr Gautam Vaddadi Northern Hospital, The Dr Siobhan Lockwood & Prof James Cameron MonashHeart (Monash Medical Centre Clayton) A/Prof James Wong Royal Melbourne Hospital, The A/Prof David Prior St Vincent’s Hospital Melbourne Dr Chris Neil Sunshine Hospital A/Prof John Amerena University Hospital, Geelong

VCOR Program Manager Ms Angela Brennan

VCOR Project Manager Dr Diem Dinh

VCOR Project Officers Ms Rita Brien Ms Harriet Carruthers Mrs Janine Doyle

VCOR Statisticians Dr Nick Andrianopoulos Mr Mark Tacey

60 VCOR | 2016 Annual Report Funding

Funding was originally sought and obtained in 2011 from Medibank Private and the Department of Health and Human Services Victoria. Additional funding from the Victorian Cardiac Clinical Network was provided in 2014, 2015 and 2016 to support ongoing operations of the PCI, Early Management of Acute STEMI (Regional Victoria) and Heart Failure data collection modules (Table 28). Considerable in-kind support from Monash University has also been provided throughout the years.

Table 28: VCOR Funding 2011 – 2016

Funding Body 2011 2012 2013 2014 2015 2016 $ $ $ $ $ $ Medibank Private $100,000 $400,000 $400,000 $300,000 - - Department of Health & Human Services $200,000 $200,000 $205,000 $509,466 $460,202 $834,815

Sub total $300,000 $600,000 $605,000 $809,466 $460,202 $834,815 Total received $3,609,483

3Correct at 1 June 2017

VCOR | 2016 Annual Report 61 References

1. Duckett, S., M. Cuddihy, and H. Newnham, Targeting Zero: Supporting the Victorian hospital system to eliminate avoidable harm and strengthen quality of care: Report of the review of hospital safety and quality assurance in Victoria . 2016: State of Victoria: Melbourne, Australia. Available at: https://www2.health.vic.gov.au/hospitals-and- health-services/quality-safety-service/hospital-safety-and-quality-review .

2. Victorian Government, Design, service and infrastructure plan for Victoria’s cardiac system. 2016: State of Victoria, Department of Health and Human Services: Melbourne, Australia. Available at: https://www2.health.vic.gov.au/hospitals-and-health-services/health-system-design-planning .

3. Australian Commission on Safety and Quality in Health Care, Strategic and Operating Principles and Technical Standards for Australian Clinical Quality Registries. 2010: Available at: http://www.safetyandquality.gov.au/wp- content/uploads/2012/03/Strategic-and-Operating-Principles-for-Australian-Clinical-Quality-Registries-AHMC- endorsed-Nov-2010.pdf .

4. Lefkovits, J., et al., Victorian Cardiac Outcomes Registry - 2013 Annual Public Report. 2013: Monash University, Department of Epidemiology and Preventive Medicine: Melbourne, Australia.

5. Gliklich, R. and N. Dreyer, Registries for Evaluating Patient Outcomes: A User’s Guide. 2nd edn. 2010: Agency for Healthcare Research and Quality: Rockville. MD, USA.

6. Andrianopoulos, N., et al., Quality control activities associated with registries in interventional cardiology and surgery . Heart, Lung & Circulation, 2011. 20(3): p. 180-6.

7. Lagerqvist, B., et al., Long-term outcomes with drug-eluting stents versus bare-metal stents in Sweden. The New England journal of medicine, 2007. 356(10): p. 1009-19.

8. Patel, M.R., et al., ACC/AATS/AHA/ASE/ASNC/SCAI/SCCT/STS 2017 Appropriate Use Criteria for Coronary Revascularization in Patients With Stable Ischemic Heart Disease: A Report of the American College of Cardiology Appropriate Use Criteria Task Force, American Association for Thoracic Surgery, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, and Society of Thoracic Surgeons . Journal of the American College of Cardiology, 2017. 69(17): p. 2212-2241.

9. Patel, M.R., et al., ACC/AATS/AHA/ASE/ASNC/SCAI/SCCT/STS 2016 Appropriate Use Criteria for coronary revascularization in patients with acute coronary syndromes. Journal of Nuclear Cardiology, 2017. 24(2): p. 439-463.

10. Eikelboom, J.W., et al., Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation, 2006. 114(8): p. 774-82.

11. Lee, P.H., et al., Left main coronary artery disease: secular trends in patient characteristics, treatments, and outcomes. Journal of the American College of Cardiology, 2016. 68(11): p. 1233-1246.

12. Jolly, S.S., et al., Thrombus aspiration in ST-Segment–Elevation Myocardial Infarction. Circulation, 2017. 135(2): p. 143-152.

13. Chew, D.P., et al., National Heart Foundation of Australia and Cardiac Society of Australia and New Zealand: Australian clinical guidelines for the management of acute coronary syndromes 2016 . Med J Aust, 2016. 205(3): p. 128-133.

14. Karrowni, W., et al., Radial versus femoral access for primary percutaneous interventions in ST-segment elevation myocardial infarction patients: a meta-analysis of randomized controlled trials . JACC: Cardiovascular Interventions, 2013. 6(8): p. 814-823.

62 VCOR | 2016 Annual Report 15. O’Gara, P.T., et al., 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Journal of the American College of Cardiology, 2013. 61(4): p. 485-510.

16. Chew, D.P., et al., Acute coronary syndrome care across Australia and New Zealand: the SNAPSHOT ACS study. Med J Aust, 2013. 199(3): p. 185-91.

17. Andrianopoulos N, C.W., Reid CM, Brennan AL, Yan BP, Yip T, Clark DJ, Ajani AE and Duffy SJ. , Australia’s First PCI Registry-Derived Logistic and Additive Risk Score Calculations Predicting Post-Procedural Adverse Outcomes. Global Heart, Supplement, 2014. 9(1): p. e308.

18. Mehran, R., et al., Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. Circulation, 2011. 123(23): p. 2736-47.

19. Williams, A., EuroQol - a new facility for the measurement of health-related quality of life. Health policy, 1990. 16(3): p. 199-208.

20. Australian Institute of Health and Welfare, Monitoring acute coronary syndrome using national hospital data: an information paper on trends and issues. Cat. no. CVD 57. 2010: AIHW: Canberra. Available at: http://www.aihw.gov.au/publication-detail/?id=10737420977 .

21. Steg, P.G., et al., ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST- segment elevation: The Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC). European Heart Journal, 2012. 33(20): p. 2569-619.

22. Darling, C.E., et al., Survival after hospital discharge for ST-segment elevation and non-ST-segment elevation acute myocardial infarction: a population-based study. Clinical Epidemiology, 2013. 5: p. 229-236.

23. Weston, C., et al., Myocardial Ischaemia National Audit Project. How the NHS cares for patients with heart attack: Annual Public Report April 2013–March 2014. 2014: University College, London, National Institute for Cardiovascular Outcomes Research.

24. Australian Institute of Health and Welfare., Australia’s health 2010 in Australia’s health series no. 12. Cat. no. AUS 122. 2010: AIHW: Canberra. Available at: http://www.aihw.gov.au/publication-detail/?id=6442468376 .

25. Heidenreich, P.A., et al., Get with the Guidelines program participation, process of care, and outcome for Medicare patients hospitalized with heart failure. Circulation: Cardiovascular Quality and Outcomes, 2012. 5(1): p. 37-43.

VCOR | 2016 Annual Report 63